EMQX Open Source Version 5

5.8.3

Release Date: 2024-12-05

Make sure to check the breaking changes and known issues before upgrading to EMQX 5.8.3.

Enhancements

Core MQTT Functionalities

• #14219 Enhanced Connection Rate Limiter for Improved System Resilience.

  • Improved system stability and responsiveness under high connection rates: Previously, when the connection rate limit was exceeded, listener acceptors would ignore new connection attempts, potentially resulting in an unrecoverable state if a large number of clients connected or reconnected frequently within a short period. Listeners now accept pending connections but immediately close them if the rate limit is reached. This reduces resource strain and improves system resilience during peak loads.

  • New listener option nolinger introduced: When set to true, a TCP-RST is sent immediately upon socket closure, helping to mitigate SYN flood attacks and further enhancing connection-handling efficiency.

  • max_connection configuration for MQTT listeners now capped by system limits: The max_connection value for MQTT listeners is now constrained by the system's limits (e.g., ulimit from the OS and node.process_limit). If configured to infinity or a value greater than the system limit, it will automatically be adjusted to match the system's maximum limit.

  • SSL listeners' ssl_options now validated before changes: Previously, invalid SSL options (such as unsupported TLS versions) could be accepted, causing client connection failures after a listener reconfiguration. With this update:

    • The node will fail to boot if a listener is configured with invalid SSL options.
    • Requests to apply invalid SSL options via the Dashboard or config API will now fail with a 400 status code.

Configuration

• #14195 Added support for client ID override.

  EMQX now provides greater flexibility by allowing custom client ID overrides using the mqtt.clientid_override={Expression} configuration. This introduces a more dynamic approach to client ID management. As part of this update, the use_userid_as_clientid and peer_cert_as_clientid options are deprecated, though they will remain available for compatibility until version 6.0.

MQTT over QUIC

• #14283 Improved QUIC transport, upgrade quicer to 0.1.9.
  • Early release of remote stream resources in the abnormal scenarios.
  • Added more troubleshooting APIs. For more details, see: https://github.com/emqx/quic/compare/0.1.6...0.1.9.

Bug Fixes

Core MQTT Functionalities

• #14201 Prevent check_gc warning from appearing when a WebSocket connection encounters a rate limit.

• #14215 Fixed an issue where calls to the retainer (via REST or CLI) would throw an exception if it was disabled.

• #14223 Ensured the WebSocket close reason is returned as an atom to avoid crashes.

• #14260 Resolved a rare race condition that could cause the connection process to crash if the CONNECT packet was not fully received before the idle timeout (default 15 seconds) expired.

• #14268 Fixed another rare race condition that could cause the WebSocket connection process to crash when the CONNECT packet was not fully received before the idle timeout expired.

• #14266 Updated emqtt from version 1.13.0 to 1.13.5. For more details, please refer to the emqtt changelog.

Durable Sessions

• #14160 Ensured that topic matching rules for durable session subscriptions are properly applied to topics starting with the $ symbol, in accordance with the MQTT specification.

REST API

• #14117 Fixed an issue in the REST API documentation where the Users endpoint was incorrectly listed as supporting Basic Authentication.

Authentication

• #14305 Removed support of hashing algorithms MD4, MD5, and RIPEMD-160 from authentication as they are not compliant with NIST Secure Hash Standard.

Data Integration

• #14172 Resolved a potential race condition where testing a connector using the HTTP API could leave lingering resources if the HTTP request timed out.

• #14178 Fixed an issue where configuration synchronization could become stuck on a particular node due to simultaneous deletion of rules across different nodes in the cluster.

• #14226 Mitigated a scenario where, under high load, a node could lose track of resource metrics (e.g., action/source) and fail to recover without a restart. Now, when restarting a resource or resetting its metrics, the system attempts to recreate the lost metrics.

  Additionally, warning logs related to metric failures, such as those for "hot-path" metrics like matched, are now throttled to prevent excessive log flooding.

• #14265 Fixed an issue where a badkey error would occur when stopping a connector if the MQTT Source action failed to subscribe successfully.

• #14296 Prevented ecpool_sup from being blocked by a slow-starting ecpool_worker.

Configuration

• #14180 Fixed an issue with variform expressions returning 'undefined' when a variable is bound to the value undefined or null. Now, an empty string is returned instead.

• #14289 Resolved a log file path issue when importing configurations from a different environment. The EMQX_LOG_DIR environment variable is set to /opt/emqx/log in Docker but /var/log/emqx/ when installed via RPM/DEB packages. Prior to this fix, log file paths (default file handler and audit handler) are environment-variable interpolated when being exported. This could cause crashes when importing configs into a different environment where the directory didn’t exist.

  With this fix, log file paths are no longer environment-variable interpolated during export. Additionally, absolute log directory paths from older versions are now converted back to environment variables if the path doesn’t exist in the new environment.

Extension

• #14243 Fixed an issue where the client.connect hook was not being triggered for some gateways.

MQTT over QUIC

• #14258 Reduced the QUIC connection shutdown timeout. Previously, QUIC connections had a 5-second timeout for graceful shutdown. If the client was unresponsive, EMQX would log warnings like:

  [warning] msg: session_stepdown_request_timeout, action: discard,

  or potentially cause a timeout on the Dashboard when attempting to disconnect the client. The timeout has now been reduced to 1 second for "kick" actions and 3 seconds for other scenarios.

5.8.2

Release Date: 2024-11-12

Make sure to check the breaking changes and known issues before upgrading to EMQX 5.8.2.

Enhancements

Core MQTT Functionalities

• #14059 Added a new configuration option for the retainer to cap message expiry intervals for retained messages. This enables garbage collection to remove messages sooner if storage is running low.

• #14072 Updated the virtual machine to use Unicode for its printable range. This improvement enhances the readability of certain binary data in messages. For instance, a binary previously displayed as <<116,101,115,116,228,184,173,230,150,135>> will now be formatted as <<"test中文"/utf8>>, providing clearer representation.

MQTT Durable Sessions

• #14130 Reduced CPU usage for idle durable sessions.

  Previously, idle durable sessions periodically woke up to refresh the list of DS streams. With this change, stream discovery is now event-based, significantly lowering CPU consumption during idle periods. Additionally, the update reduces the delay in notifying sessions of new streams, effectively eliminating the long-tail latency in end-to-end processing.

REST API

• #13889 Enhanced the performance of the /api/v5/monitor_current and /api/v5/metrics APIs.

  Previously, these APIs queried clustered nodes sequentially in a loop. Now, the queries are sent in parallel, reducing response time. The latency is now primarily dependent on the slowest node in the cluster.

  Additionally, a node parameter was added to the /api/v5/monitor_current API, allowing targeted queries to a single node instead of the entire cluster. For instance, using ?aggregate=false&node=emqx@node1.domain.name will return data exclusively for the specified node.

EMQX Clustering

• #13903 Added logs to inform the user when a replicant node cannot find a core node with the same release version as its own.

Security

• #13923 Added zone support in authentication, authorization, and mountpoint templates.

  Previously, to reference a client's zone in authentication or authorization rules, users needed to access it through client_attrs. Now, the ${zone} placeholder can be used directly in these templates, simplifying rule creation and enabling zone-specific configurations.

  For example, the following ACL rule uses ${zone} to dynamically apply permissions based on a client’s assigned zone: {allow, all, all, ["${zone}/${username}/#"]}.

• #14102 Added support for SSL private key passphrase from a secret file.

  EMQX can now read the passphrase from a secret file if password is configured as ...ssl_options.password = "file://{path-to-secret-file}".

Data Integration

• #14065 Added a new queuing_bytes metric for data integration. This metric shows the RAM and/or disk resources consumed by buffering for a specific action.

Observability

• #14096 Exposed emqx_conf_sync_txid as a Prometheus metric, allowing for monitoring the configuration file synchronization status of each node in the cluster.

MQTT over QUIC

• #13814 Connection Scope Keepalive for MQTT over QUIC Multi-Stream:

  Introduced a new feature to keep MQTT connections alive when data streams remain active, even if the control stream is idle.

  Previously, clients were required to send MQTT.PINGREQ on idle control streams to keep the connection alive. Now, a shared state tracks activity across all streams for each connection. This shared state is used to determine if the connection is still alive, reducing the risk of keepalive timeouts due to Head-of-Line (HOL) blocking.

• #14112 Added support ssl_options.hibernate_after in QUIC listener to reduce memory footprint of QUIC transport.

Bug Fixes

Core MQTT Functionality

• #13931 Updated the gen_rpc library to version 3.4.1, which includes a fix to prevent client socket initialization errors from escalating to the node level on the server side.
• #13969 Optimized the periodic cleanup of expired retained messages to ensure efficient resource usage, particularly in cases with a large volume of expired messages.
• #14068 Added the handle_frame_error/2 callback to all gateway implementation modules to handle message parsing errors.
• #14037 Improved the internal database bootstrap process to better tolerate temporary unavailability of peer nodes, particularly when a new node joins an existing cluster.
• #14116 Fixed an issue where the default configuration for the retainer was generated incorrectly after joining a cluster.

MQTT Durable Sessions

• #14042 Fix crash in the durable session after updates to subscription parameters (such as QoS, no_local, upgrade_qos, ...).
• #14052 Corrected memory usage reporting from cgroups when in use.
• #14055 Updated the /clients_v2 API to properly respect all filtering arguments when querying offline clients with durable sessions. Previously, only the username filter was applied, while other filtering arguments were ignored.
• #14151 Fixed handling of the conn_state filter in the /clients_v2 API for offline clients with durable sessions. Previously, these clients could be incorrectly selected with conn_state=connected.
• #14057 Resolved a compatibility issue that prevented the Messages DS database from starting due to a slightly different database configuration schema. This issue occurred when upgrading EMQX from version 5.7.x with session durability enabled.

REST API

• #14023 Fixed an issue with the GET /monitor HTTP API where returned values could appear higher than actual values, depending on the requested time window. For data points within a 1-hour window, this distortion is only visual on the Dashboard. However, for data points older than 1 hour, the data distortion is permanent.

  The affected metrics include:

  • disconnected_durable_sessions
  • subscriptions_durable
  • subscriptions
  • topics
  • connections
  • live_connections

EMQX Clustering

• #13996 Fixed an intermittent crash occurring when using emqx conf fix to resolve configuration discrepancies, particularly if a configuration key was missing on one of the nodes.

Security

• #13922 Updated the CRL (Certificate Revocation List) cache to use the full Distribution Point (DP) URL as the cache key. Previously, only the path part of the URL was used, causing conflicts when multiple DPs shared the same path.
• #13924 Fixed an issue where JWK keys could leak into debug logs upon JWT authentication failure.

Data Integration

• #13916 Fixed an issue where the parent metric failed was not incremented when a rule’s failed.no_result or failed.exception metrics were updated.

• #14001 Resolved a race condition where a resource (such as a connector, action, source, authentication, or authorization) could falsely report a connected, healthy channel after a brief disconnection. This issue could result in excessive action_not_found log entries when the race condition occurred.

• #13913 Fixed an issue with the actions and source HTTP APIs where a 500 status code would be returned if a timeout occurred while attempting to update or delete a resource.

• #14101 Resolved an issue where deleting a resource would fail if a source and an action were both created with the same name.

Observability

• #13909 Fixed log formatting for cases where the payload cannot be displayed as readable UTF-8 Unicode characters.

• #14061 Improved log information when emqx_cm:request_stepdown/3 fails.

  In scenarios where a client channel needs to terminate another channel with the same ClientID, a race condition may occur if the target channel has already been closed or terminated. In such cases, error logs and stack traces that provide no useful information will no longer be generated.

• #14070 Removed the connector's state from error and warning logs due to its potential length. For issue analysis, the connector's state can now be accessed through emqx_resource:list_instances_verbose/0. Below is an example of a log entry before this change:

  pid: <0.43914.0>, connector: connector:sqlserver:connector-05a2e105, reason: [Microsoft][ODBC Driver 17 for SQL Server][SQL Server]Argument data type varchar is invalid for argument 2 of dateadd function. SQLSTATE IS: 42000, state: {"resource_opts":{"start_timeout":5000,"start_after_created":true,"health_check_interval":15000},"pool_name":"connector:sqlserver:connector-05a2e105","installed_channels":{"action:sqlserver:action-4b033621:connector:sqlserver:connector-05a2e105":{"sql_templates":{"batch_insert_temp":{"send_message":{"batch_insert_tks":["{str,<<\" ( \">>}","{var,[<<\"messageId\">>]}","{str,<<\", \">>}","{var,[<<\"measurement\">>]}","{str,<<\", \">>}","{var,[<<\"Analog_IN_Fault_1\">>]}","{str,<<\", \">>}","{var,[<<\"Analog_IN_Fault_2\">>]}","{str,<<\", \">>}","{var,[<<\"Analog_IN_Fault_3\">>]}","{str,<<\", \">>}","{var,[<<\"Analog_IN_Fault_4\">>]}","{str,<<\", \">>}","{var,[<<\"Analog_IN_PV_1\">>]}","{str,<<\", \">>}","{var,[<<\"Analog_IN_PV_2\">>]}","{str,<<\", \">>}","{var,[<<\"Analog_IN_PV_3\">>]}","{str,<<\", \">>}","{var,[<<\"Analog_IN_PV_4\">>]}","{str,<<\", DATEADD(MS, \">>}","{var,[<<\"ms_shift\">>]}","{str,<<\", DATEADD(S, \">>}","{var,[<<\"s_shift\">>]}","{str,<<\", '19700101 00:00:00:000') ))\">>}"],"batch_insert_part":"insert into TransactionLog(MessageId, Measurement, Fault1, Fault2, Fault3, Fault4, Value1, Value2, Value3, Value4, DateStamp) \r\n"}}}}}},msg: invalid_request

• #14099 Removed an error-level log entry that was triggered when validation of UTF-8 strings in MQTT messages failed.

• #14091 Implemented a fix to remove function_clause from log messages when users provide unsupported write syntax.

  Example of unsupported syntax:

  weather,location=us-midwest,season=summer temperature=82 ${timestamp}u

  Before this fix, the error log would contain the function_clause error, as shown:

  pid: <0.558392.0>, info: {"stacktrace":["{emqx_bridge_influxdb_connector,parse_timestamp,[[1719350482910000000,<<\"u\">>]],[{file,\"emqx_bridge_influxdb_connector.erl\"},{line,692}]}", ...], ..., "error":"{error,function_clause}"}, tag: ERROR, msg: resource_exception

  This change improves log clarity by omitting function_clause in cases of syntax errors.

5.8.1

Release Date: 2024-10-14

Make sure to check the breaking changes and known issues before upgrading to EMQX 5.8.1.

Important Changes

• #13956 Updated the gen_rpc library to version 3.4.1, which includes a node crash issue. Previously, if a node is force shutdown down while RPC channels are being established, it may cause a cluster peer node to crash.

Enhancements

Core MQTT Functionalities

• #13525 Added new configuration item shared_subscription_initial_sticky_pick to specify the strategy for making the initial pick when shared_subscription_strategy is set to sticky.

• #13942 The HTTP client now automatically reconnects if no activity is detected for 10 seconds after the latest request has expired. Previously, it would wait indefinitely for a server response, causing timeouts if the server dropped requests.

  This change impacts below components.

  • HTTP authentication
  • HTTP authorization
  • Webhook (HTTP connector)

Authentication and Authorization

• #13863 EMQX now supports ${cert_common_name} placeholder in topic name templates for raw ACL rules.

• #13792 The banned-clients API GET /banned supports querying the rules using filters in the query string.

  The available filters are:

  • clientid
  • username
  • peerhost
  • like_clientid
  • like_username
  • like_peerhost
  • like_peerhost_net

  When adding a new banned client entry, the default expiration time for entries without the until parameter specified has been changed from 1 year to infinite.

Rule Engine

• #13773 Disabled rule actions now do not trigger out_of_service warnings.

  Previously, if an action is disabled, there would be a warning log with msg: out_of_service, and the actions.failed counter was incremented for the rule.

  After this enhancement, disabled action will result in a debug level log with msg: discarded, and the newly introduced counter actions.discarded will be incremented.

MQTT over QUIC

• #13814 Connection Scope Keepalive for MQTT over QUIC Multi-Stream:

  This update introduces a new feature to maintain MQTT connections over QUIC multi-streams, even when the control stream is idle but other data streams are active.

  Previously, clients had to send MQTT.PINGREQ on idle control streams to keep the connection alive. Now, a shared state is maintained for each connection, monitoring activity across all streams. This shared state helps determine if the connection is still active, reducing the risk of keepalive timeouts caused by Head-of-Line (HOL) blocking and improving overall connection stability.

Bug Fixes

Core MQTT Functions

• #13702 Clean up the corresponding exclusive subscriptions when a node goes down.

• #13708 Fixed an issue which may cause shared subscription 'sticky' strategy to degrade to 'random'.

• #13733 Made cacertfile optional when configuring https listener from emqx ctl conf load command.

• #13742 Fixed when subscribing with + as the first level, or # as a wildcard, retained messages with topics starting with $ are incorrectly received.

• #13754 Fixed an issue when websocket connection would break consistently on its own.

• #13756 Introduced more randomness to broker assigned client IDs.

• #13790 The default heartbeat interval for the MQTT connector has been reduced from 300 seconds to 160 seconds.

  This change helps maintain the underlying TCP connection by preventing timeouts due to the idle limits imposed by load balancers or firewalls, which typically range from 3 to 5 minutes depending on the cloud provider.

• #13832 Fixed that the Publish endpoint would have a 500 error when persistent session were enabled.

• #13842 Fixed a UTF-8 string validation exception.

Upgrade and Migration

• #13731 Resolved an issue that prevented clusters running on EMQX 5.4.0 from upgrading to EMQX 5.8.0. This fix introduces a migration procedure to update specific internal database tables created in version 5.4.0 to align with the new schema.

5.8.0

Release Date: 2024-08-28

Enhancements

Core MQTT Functionality

• #13009 Updated the log level for message receiving pause due to rate limiting from debug to warning. The log message socket_receive_paused_by_rate_limit is throttled to avoid excessive logging.

Authentication and Authorization

• #12418 Enhanced JWT authentication to support claims verification using a list of objects:

  [
    {
      name = "claim_name",
      value = "${username}"
    },
    ...
  ]

  Expected values are now treated as templates, consistent with other authenticators, allowing for arbitrary expressions such as ${username} and ${clientid}. Previousy, only fixed "${username}" "${clientid}" values were supported for interpolation.

  Improved the documentation for the verify_claims parameter.

• #13229 Added support for ${cert_pem} placeholder in authentication templates.

• #13534 Added trace logging to indicate when the superuser bypasses the authorization check.

Data Integrations

• #13144 Changed the log level to warning and added throttling for the log message data_bridge_buffer_overflow when bridge buffers overflow and messages are dropped. Previously, these events were logged at the info level and were not visible with the default log settings.

• #13492 Enhanced the GET /connectors and GET /connectors/:id APIs to include lists of actions and sources that depend on a specific connector. Additionally, the GET /actions, GET /sources, GET /actions/:id, and GET /sources/:id APIs now return the list of rules associated with a specific action or source.

• #13505 Added the ability to filter rules in the HTTP API based on the IDs of data integration actions or sources used.

• #13506 Introduced the peername field to all rule engine events that already include the peerhost field. The peername field is a string formatted as IP:PORT.

• #13516 Added a direct_dispatch argument to the republish action.

  When direct_dispatch is set to true (or rendered as true from template) the message is dispatched directly to subscribers. This feature helps prevent the triggering of additional rules or the recursive activation of the same rule.

• #13573 Introduced client_attrs to the SQL context for client connectivity events and the message publish event. Users can now access client attributes within rule SQL statements, such as SELECT client_attrs.attr1 AS attribute1, and utilize ${attribute1} in data integration actions.

• #13640 Added two new SQL functions for rules: coalesce/2 and coalesce_ne/2.

  These functions simplify handling null values in rule SQL expressions. For instance, instead of using:

  SELECT
    CASE
      WHEN is_null(payload.path.to.value) THEN
        0
      ELSE
        payload.path.to.value
    END AS my_value

  you can now write a more concise expression: SELECT coalesce(payload.path.to.value, 0) AS my_value.

Operations

• #13202 Introduced the emqx ctl conf cluster_sync fix command to address cluster configuration inconsistencies. This command synchronizes the configuration of all nodes with the configuration of the node that has the highest tnx_id, ensuring consistency across the cluster.

• #13250 Added a new value for cluster.discovery_strategy: singleton. By choosing this option, there will be effectively no clustering, and the node will reject connection attempts to and from other nodes.

• #13370 Added a new version of wildcard_optimized storage layout for durable storage, offering the following improvements:

  • The new layout does not have an inherent latency.

  • MQTT messages are serialized into a more space-efficient format.

• #13524 Added the emqx ctl exclusive CLI interface to manage exclusive topics more effectively. It allows administrators to better manage and troubleshoot exclusive topic subscriptions, ensuring that subscription states are accurately reflected and preventing unexpected failures.

• #13597 Added thin wrapper functions for plugins to store and manage the certificate files used by the plugins themselves. This fix prevents plugin certificates from being inadvertently deleted by the certificate garbage collection (GC) function.

• #13626 Added a new command emqx ctl listeners enable <Identifier> <Bool> to enable/disable a listener.

• #13493 Upgraded the RPC library gen_rpc to version 3.4.0. This update changes the default RPC server socket option from true to active-100, which introduces back-pressure to peer nodes when the RPC server experiences heavy load.

• #13665 Added a new metric emqx_actions_count to the prometheus endpoint. It contains the number of all actions added by all rules, including Republish actions and Console Output actions.

Bug Fixes

Core MQTT Functionality

• #12944 Fixed an issue that caused a crash when clients with non-UTF8 client IDs attempted to connect with strict_mode=false.

• #13006 Improved the validation of retained, delayed, and taken-over session messages to ensure they comply with banned client ID rules implemented through regular expression matching. Previously, certain messages, such as those delayed due to network issues or taken over by another session, could bypass the client ID bans set by regular expressions.

Authentication and Authorization

• #13024 Added a default ACL deny rule to reject subscriptions to the +/# topic pattern. Since EMQX by default rejects subscriptions to # topic, for completeness, it should reject +/# as well.

• #13040 Improved HTTP authentication:

  • Improved error logging for cases where the HTTP Content-Type header is missing or unrecognized, providing more detailed information.
  • Fixed an issue causing double encoding of query parameters in authentication HTTP requests
  • Enhanced error messages when a POST method with a JSON content type is configured for authentication requests but the JSON template fails to render into valid JSON. This can occur, for example, when a template contains a placeholder like ${password} but receives a non-UTF8 password input, leading to better transparency and easier debugging for such scenarios.

• #13196 Added a limit to the built-in authorization database, restricting the number of Access Control List (ACL) rules per client or user to a default of 100.

• #13584 Fixed an issue with creating HTTP authorization that resulted in errors when the HTTP header list was empty.

• #13618 Improved the type specifications for the authorization/sources endpoint to provide clearer and more concise error messages.

• #13624 Fixed an issue in the built-in authorizer where updating rules for a client or user could result in the total number of rules exceeding the max_rules limit.

• #13678 Made the deletion of an authenticator in the chain an idempotent operation, ensuring that deleting a non-existing authenticator always succeeds.

Data Integrations

• #13207 Improved the republish rule engine action to accurately reflect the success and failure of message publishing. Previously, the success metrics were incremented even when the republish action failed to deliver the message to any subscribers. Now, if the action detects that a message fails to reach any subscriber, the failure metrics are correctly incremented.

• #13425 Improved the MQTT connector error log messages to provide clearer and more detailed information.

• #13589 Fixed an issue where creating a rule with a string "null" for ID via the HTTP API was allowed, which could lead to an inconsistent configuration.

Operations

• #13078 Improved validation and error handling in the EMQX Management API to ensure that requests with a JSON body include the Content-Type: application/json header. If the header is missing for APIs that expect JSON input, the server now correctly responds with a 415 Unsupported Media Type status code instead of 400 Bad Request.

• #13225 Enhanced security in authentication and authorization APIs by redacting sensitive data such as passwords. Previously, the APIs could return the original password values in responses. With this update, sensitive information is replaced with ****** to prevent accidental exposure and protect user credentials.

Gateways

• #13607 Fixed an issue where the QoS level for CoAP subscriptions displayed through the API did not match the actual QoS level being used. This discrepancy could cause confusion as successful subscriptions were not accurately reflected on the Dashboard.

5.7.2

Release Date: 2024-08-07

Enhancements

• #13317 Added a new per-authorization source metric type: ignore. This metric increments when an authorization source attempts to authorize a request but encounters scenarios where the authorizer is not applicable or encounters an error, resulting in an undecidable outcome.

• #13336 Added functionality to initialize authentication data in the built-in database of an empty EMQX node or cluster using a bootstrap file in CSV or JSON format. This feature introduces new configuration entries, bootstrap_file and bootstrap_type.

• #13348 Added a new field payload_encode in the log configuration to determine the format of the payload in the log data.

• #13436 Added the option to add custom request headers to JWKS requests.

• #13507 Introduced a new built-in function getenv in the rule engine and variform expression to facilitate access to environment variables. This function adheres to the following constraints:

  • Prefix EMQXVAR_ is added before reading from OS environment variables. For example, getenv('FOO_BAR') is to read EMQXVAR_FOO_BAR.
  • These values are immutable once loaded from the OS environment.

• #13521 Resolved an issue where LDAP query timeouts could cause the underlying connection to become unusable, potentially causing subsequent queries to return outdated results. The fix ensures the system reconnects automatically in case of a timeout.

• #13528 Applied log throttling for the event of unrecoverable errors in data integrations.

• #13548 EMQX now can optionally invoke the on_config_changed/2 callback function when the plugin configuration is updated via the REST API. This callback function is assumed to be exported by the <PluginName>_app module. For example, if the plugin name and version are my_plugin-1.0.0, then the callback function is assumed to be my_plugin_app:on_config_changed/2.

• #13386 Added support for initializing a list of banned clients on an empty EMQX node or cluster with a bootstrap file in CSV format. The corresponding config entry to specify the file path is banned.bootstrap_file. This file is a CSV file with , as its delimiter. The first line of this file must be a header line. All valid headers are listed here:

  • as :: required
  • who :: required
  • by :: optional
  • reason :: optional
  • at :: optional
  • until :: optional

  See the Configuration Manual for details on each field.

  Each row in the rest of this file must contain the same number of columns as the header line, and the column can be omitted then its value is undefined.

Bug Fixes

• #13222 Resolved issues with flags checking and error handling associated with the Will message in the CONNECT packet. For detailed specifications, refer to:

  • MQTT-v3.1.1-[MQTT-3.1.2-13], MQTT-v5.0-[MQTT-3.1.2-11]
  • MQTT-v3.1.1-[MQTT-3.1.2-14], MQTT-v5.0-[MQTT-3.1.2-12]
  • MQTT-v3.1.1-[MQTT-3.1.2-15], MQTT-v5.0-[MQTT-3.1.2-13]

• #13307 Updated ekka library to version 0.19.5. This version of ekka utilizes mria 0.8.8, enhancing auto-heal functionality. Previously, the auto-heal worked only when all core nodes were reachable. This update allows to apply auto-heal once the majority of core nodes are alive. For details, refer to the Mria PR.

• #13334 Implemented strict mode checking for the PasswordFlag in the MQTT v3.1.1 CONNECT packet to align with protocol specifications.

  Note: To ensure bug-to-bug compatibility, this check is performed only in strict mode.

• #13344 Resolved an issue where the POST /clients/:clientid/subscribe/bulk API would not function correctly if the node receiving the API request did not maintain the connection to the specified clientid.

• #13358 Fixed an issue when the reason in the authn_complete_event event was incorrectly displayed.

• #13375 The value infinity has been added as default value to the listener configuration fields max_conn_rate, messages_rate, and bytes_rate.

• #13382 Updated the emqtt library to version 0.4.14, which resolves an issue preventing emqtt_pools from reusing pools that are in an inconsistent state.

• #13389 Fixed an issue where the Derived Key Length for pbkdf2 could be set to a negative integer.

• #13389 Fixed an issue where topics in the authorization rules might be parsed incorrectly.

• #13393 Fixed an issue where plugin applications failed to restart after a node joined a cluster, resulting in hooks not being properly installed and causing inconsistent states.

• #13398 Fixed an issue where ACL rules were incorrectly cleared when reloading the built-in database for authorization using the command line.

• #13403 Addressed a security issue where environment variable configuration overrides were inadvertently logging passwords. This fix ensures that passwords present in environment variables are not logged.

• #13408 Resolved a function_clause crash triggered by authentication attempts with invalid salt or password types. This fix enhances error handling to better manage authentication failures involving incorrect salt or password types.

• #13419 Resolved an issue where crash log messages from the /configs API were displaying garbled hints. This fix ensures that log messages related to API calls are clear and understandable.

• #13422 Fixed an issue where the option force_shutdown.max_heap_size could not be set to 0 to disable this tuning.

• #13442 Fixed an issue where the health check interval configuration for actions/sources was not being respected. Previously, EMQX ignored the specified health check interval for actions and used the connector's interval instead. The fix ensures that EMQX now correctly uses the health check interval configured for actions/sources, allowing for independent and accurate health monitoring frequencies.

• #13503 Fixed an issue where connectors did not adhere to the configured health check interval upon initial startup, requiring an update or restart to apply the correct interval.

• #13515 Fixed an issue where the same client could not subscribe to the same exclusive topic when the node was down for some reason.

• #13527 Fixed an issue in the Rule Engine where executing a SQL test for the Message Publish event would consistently return no results when a $bridges/... source was included in the FROM clause.

• #13541 Fixed an issue where disabling CRL checks for a listener required a listener restart to take effect.

• #13552 Added a startup timeout limit for EMQX plugins with a default timeout of 10 seconds. Before this update, problematic plugins could cause runtime errors during startup, leading to potential issues where the main startup process might hang when EMQX is stopped and restarted.

5.7.1

Release Date: 2024-06-26

Enhancements

• #12983 Add new rule engine event $events/client_check_authn_complete for authentication completion event.

• #13180 Improved client message handling performance when EMQX is running on Erlang/OTP 26 and increased message throughput by 10% in fan-in mode.

• #13191 Upgraded EMQX Docker images to run on Erlang/OTP 26.

  EMQX had been running on Erlang/OTP 26 since v5.5 except for docker images which were on Erlang/OTP 25. Now all releases are on Erlang/OTP 26.

• #13242 Significantly increased the startup speed of EMQX dashboard listener.

Bug Fixes

• #13156 Resolved an issue where the Dashboard Monitoring pages would crash following the update to EMQX v5.7.0.

• #13164 Fixed HTTP authorization request body encoding.

  Before this fix, the HTTP authorization request body encoding format was taken from the accept header. The fix is to respect the content-type header instead. Also added access templating variable for v4 compatibility. The access code of SUBSCRIBE action is 1 and PUBLISH action is 2.

• #13238 Improved the logged error messages when an HTTP authorization request with an unsupported content-type header is returned.

• #13258 Fix an issue where the MQTT-SN gateway would not restart correctly due to incorrect startup order of gateway dependencies.

• #13273 Fixed and improved handling of URIs in several configurations. The fix includes the following improvement details:

  • Authentication and authorization configurations: Corrected a previous error where valid pathless URIs such as https://example.com?q=x were mistakenly rejected. These URIs are now properly recognized as valid.
  • Connector configurations: Enhanced checks to ensure that URIs with potentially problematic components, such as user info or fragment parts, are no longer erroneously accepted.

• #13276 Fixed an issue in the durable message storage mechanism where parts of the internal storage state were not correctly persisted during the setup of new storage generations. The concept of "generation" is used internally and is crucial for managing message expiration and cleanup. This could have manifested as messages being lost after a restart of EMQX.

• #13291 Fixed an issue where durable storage sites that were down being reported as up.

• #13290 Fixed an issue where the command $ bin/emqx ctl rules show rule_0hyd would produce no output when used to display rules with a data integration action attached.

• #13293 Improved the restoration process from data backups by automating the re-indexing of imported retained messages. Previously, re-indexing required manual intervention using the emqx ctl retainer reindex start CLI command after importing a data backup file.

  This fix also extended the functionality to allow exporting retained messages to a backup file when the retainer.backend.storage_type is configured as ram. Previously, only setups with disc as the storage type supported exporting retained messages.

• #13140 Fixed an issue that caused text traces for the republish action to crash and not display correctly.

• #13148 Fixed an issue where a 500 HTTP status code could be returned by /connectors/:connector-id/start when there is a timeout waiting for the resource to be connected.

• #13181 EMQX now forcefully shut down the connector process when attempting to stop a connector, if such operation times out. This fix also improved the clarity of error messages when disabling an action or source fails due to an unresponsive underlying connector.

• #13216 Respect clientid_prefix config for MQTT bridges. Since EMQX v5.4.1, the MQTT client IDs are restricted to a maximum of 23 bytes. Previously, the system factored the clientid_prefix into the hash of the original, longer client ID, affecting the final shortened ID. The fix includes the following change details:

  • Without Prefix: The behavior remains unchanged. EMQX hashes the long client IDs (exceeding 23 bytes) to fit within the 23-byte limit.
  • With Prefix:
    • Prefix ≤ 19 bytes: The prefix is retained, and the remaining portion of the client ID is hashed into a 4-byte space, ensuring the total length does not exceed 23 bytes.
    • Prefix ≥ 20 bytes: EMQX will not attempt to shorten the client ID, fully preserving the configured prefix regardless of length.

5.7.0

Release Date: 2024-05-27

Enhancements

Security

#12947 For JWT authentication, support new disconnect_after_expire option. When enabled, the client will be disconnected after the JWT token expires.

Note: This is a breaking change. This option is enabled by default, so the default behavior is changed. Previously, the clients with actual JWTs could connect to the broker and stay connected even after the JWT token expired. Now, the client will be disconnected after the JWT token expires. To preserve the previous behavior, set disconnect_after_expire to false.

Data Processing and Integration

#12671 An unescape function has been added to the rule engine SQL language to handle the expansion of escape sequences in strings. This addition has been done because string literals in the SQL language don't support any escape codes (e.g., \n and \t). This enhancement allows for more flexible string manipulation within SQL expressions.

Extensibility

• #12872 Implemented the Client Attributes feature. It allows setting additional properties for each client using key-value pairs. Property values can be generated from MQTT client connection information (such as username, client ID, TLS certificate) or set from data accompanying successful authentication returns. Properties can be used in EMQX for authentication, authorization, data integration, and MQTT extension functions. Compared to using static properties like client ID directly, client properties offer greater flexibility in various business scenarios, simplifying the development process and enhancing adaptability and efficiency in development work. Initialization of client_attrs The client_attrs fields can be initially populated from one of the following clientinfo fields:

  • cn: The common name from the TLS client's certificate.
  • dn: The distinguished name from the TLS client's certificate, that is, the certificate "Subject".
  • clientid: The MQTT client ID provided by the client.
  • username: The username provided by the client.
  • user_property: Extract a property value from 'User-Property' of the MQTT CONNECT packet.

  Extension through Authentication Responses Additional attributes may be merged into client_attrs from authentication responses. Supported authentication backends include:

  • HTTP: Attributes can be included in the JSON object of the HTTP response body through a client_attrs field.
  • JWT: Attributes can be included via a client_attrs claim within the JWT.

  Usage in Authentication and Authorization If client_attrs is initialized before authentication, it can be used in external authentication requests. For instance, ${client_attrs.property1} can be used within request templates directed at an HTTP server for authenticity validation.

  • The client_attrs can be utilized in authorization configurations or request templates, enhancing flexibility and control. Examples include: In acl.conf, use {allow, all, all, ["${client_attrs.namespace}/#"]} to apply permissions based on the namespace attribute.
  • In other authorization backends, ${client_attrs.namespace} can be used within request templates to dynamically include client attributes.

  For more information about the Client Attributes feature, see Client Attributes.

• #12910 Added plugin configuration management and schema validation. For EMQX enterprise edition, one can also annotate the schema with metadata to facilitate UI rendering in the Dashboard. See more details in the plugin template and plugin documentation.

Operations and Management

• #12923 Provided more specific error when importing wrong format into builtin authenticate database.

• #12940 Added ignore_readonly argument to PUT /configs API. Before this change, EMQX would return 400 (BAD_REQUEST) if the raw config included read-only root keys (cluster, rpc, and node). After this enhancement it can be called as PUT /configs?ignore_readonly=true, EMQX will in this case ignore readonly root config keys, and apply the rest. For observability purposes, an info level message is logged if any readonly keys are dropped. Also fixed an exception when config has bad HOCON syntax (returns 500). Now bad syntax will cause the API to return 400 (BAD_REQUEST).

• #12957 Started building packages for macOS 14 (Apple Silicon) and Ubuntu 24.04 Noble Numbat (LTS).

Bug Fixes

Security

• #12887 Fixed MQTT enhanced auth with sasl scram.

• #12962 TLS clients can now verify server hostname against wildcard certificate. For example, if a certificate is issued for host *.example.com, TLS clients is able to verify server hostnames like srv1.example.com.

MQTT

• #12996 Fixed process leak in emqx_retainer application. Previously, client disconnection while receiving retained messages could cause a process leak.

Data Processing and Integration

• #12653 The rule engine function bin2hexstr now supports bitstring inputs with a bit size that is not divisible by 8. Such bitstrings can be returned by the rule engine function subbits.

• #12657 The rule engine SQL-based language previously did not allow putting any expressions as array elements in array literals (only constants and variable references were allowed). This has now been fixed so that one can use any expressions as array elements. The following is now permitted, for example:

  select
  [21 + 21, abs(-abs(-2)), [1 + 1], 4] as my_array
  from "t/#"

• #12932 Previously, if a HTTP action request received a 503 (Service Unavailable) status, it was marked as a failure and the request was not retried. This has now been fixed so that the request is retried a configurable number of times.

• #12948 Fixed an issue where sensitive HTTP header values like Authorization would be substituted by ****** after updating a connector.

• #13118 Fix a performance issue in the rule engine template rendering.

Observability

• #12765 Make sure stats subscribers.count subscribers.max contains shared-subscribers. It only contains non-shared subscribers previously.

Operations and Management

• #12812 Made resource health checks non-blocking operations. This means that operations such as updating or removing a resource won't be blocked by a lengthy running health check.

• #12830 Made channel (action/source) health checks non-blocking operations. This means that operations such as updating or removing an action/source data integration won't be blocked by a lengthy running health check.

• #12993 Fixed listener config update API when handling an unknown zone. Before this fix, when a listener config is updated with an unknown zone, for example {"zone": "unknown"}, the change would be accepted, causing all clients to crash whens connected. After this fix, updating the listener with an unknown zone name will get a "Bad request" response.

• #13012 The MQTT listerners config option access_rules has been improved in the following ways:

  • The listener no longer crash with an incomprehensible error message if a non-valid access rule is configured. Instead a configuration error is generated.
  • One can now add several rules in a single string by separating them by comma (for example, "allow 10.0.1.0/24, deny all").

• #13041 Improved HTTP authentication error log message. If HTTP content-type header is missing for POST method, it now emits a meaningful error message instead of a less readable exception with stack trace.

• #13077 This fix makes EMQX only read action configurations from the global configuration when the connector starts or restarts, and instead stores the latest configurations for the actions in the connector. Previously, updates to action configurations would sometimes not take effect without disabling and enabling the action. This means that an action could sometimes run with the old (previous) configuration even though it would look like the action configuration has been updated successfully.

• #13090 Attempting to start an action or source whose connector is disabled will no longer attempt to start the connector itself.

Gateways

• #12909 Fixed UDP listener process handling on errors or closure, The fix ensures the UDP listener is cleanly stopped and restarted as needed if these error conditions occur.

• #13001 Fixed an issue where the syskeeper forwarder would never reconnect when the connection was lost.

• #13010 Fixed the issue where the JT/T 808 gateway could not correctly reply to the REGISTER_ACK message when requesting authentication from the registration service failed.

5.6.1

Release Date: 2024-04-18

Bug Fixes

• #12759 EMQX now automatically removes invalid backup files that fail during upload due to schema validation errors. This fix ensures that only valid configuration files are displayed and stored, enhancing system reliability.

• #12766 Renamed message_queue_too_long error reason to mailbox_overflow

  mailbox_overflow. The latter is consistent with the corresponding config parameter: force_shutdown.max_mailbox_size.

• #12773 Upgraded HTTP client libraries.

  The HTTP client library (gun-1.3) incorrectly appended a :portnumber suffix to the Host header for standard ports (http on port 80, https on port 443). This could cause compatibility issues with servers or gateways performing strict Host header checks (e.g., AWS Lambda, Alibaba Cloud HTTP gateways), leading to errors such as InvalidCustomDomain.NotFound or "The specified CustomDomain does not exist."

• #12802 Improved how EMQX handles node removal from clusters via the emqx ctl cluster leave command. Previously, nodes could unintentionally rejoin the same cluster (unless it was stopped) if the configured cluster discovery_strategy was not manual. With the latest update, executing the cluster leave command now automatically disables cluster discovery for the node, preventing it from rejoining. To re-enable cluster discovery, use the emqx ctl discovery enable command or simply restart the node.

• #12814 Improved error handling for the /clients/{clientid}/mqueue_messages and /clients/{clientid}/inflight_messages APIs in EMQX. These updates address:

  • Internal Timeout: If EMQX fails to retrieve the list of Inflight or Mqueue messages within the default 5-second timeout, likely under heavy system load, the API will return 500 error with the response {"code":"INTERNAL_ERROR","message":"timeout"}, and log additional details for troubleshooting.
  • Client Shutdown: Should the client connection be terminated during an API call, the API now returns a 404 error, with the response {"code": "CLIENT_SHUTDOWN", "message": "Client connection has been shutdown"}. This ensures clearer feedback when client connections are interrupted.

• #12824 Updated the statistics metrics subscribers.count and subscribers.max to include shared subscribers. Previously, these metrics accounted only for non-shared subscribers.

• #12826 Fixed issues related to the import functionality of source data integrations and retained messages in EMQX. Before this update:

  • The data integration sources specified in backup files were not being imported. This included configurations under the sources.mqtt category with specific connectors and parameters such as QoS and topics.
  • Importing the mnesia table for retained messages was not supported.

• #12843 Fixed cluster_rpc_commit transaction ID cleanup procedure on replicator nodes after executing the emqx ctl cluster leave command. Previously, failing to properly clear these transaction IDs impeded configuration updates on the core node.

• #12885 Fixed an issue in EMQX where users were unable to view "Retained Messages" under the "Monitoring" menu in the Dashboard.

  The "Retained messages" backend API uses the qlc library. This problem was due to a permission issue where the qlc library's file_sorter function tried to use a non-writable directory, /opt/emqx, to store temporary files, resulting from recent changes in directory ownership permissions in Docker deployments.

  This update modifies the ownership settings of the /opt/emqx directory to emqx:emqx, ensuring that all necessary operations, including retained messages retrieval, can proceed without access errors.

5.6.0

Release Date: 2024-03-28

Enhancements

• #12251 Optimized the performance of the RocksDB-based persistent sessions, achieving a reduction in RAM usage and database request frequency. Key improvements include:

  • Introduced dirty session state to avoid frequent mria transactions.
  • Introduced an intermediate buffer for the persistent messages.
  • Used separate tracks of PacketIds for QoS1 and QoS2 messages.
  • Limited the number of continuous ranges of inflight messages to 1 per stream.

• #12326 Enhanced session tracking with registration history. EMQX now has the capability to monitor the history of session registrations, including those that have expired. By configuring broker.session_history_retain, EMQX retains records of expired sessions for a specified duration.

  • Session count API: Use the API GET /api/v5/sessions_count?since=1705682238 to obtain a count of sessions across the cluster that remained active since the given UNIX epoch timestamp (with seconds precision). This enhancement aids in analyzing session activity over time.

  • Metrics expansion with cluster sessions gauge: A new gauge metric, cluster_sessions, is added to better track the number of sessions within the cluster. This metric is also integrated into Prometheus for easy monitoring:

    # TYPE emqx_cluster_sessions_count gauge
    emqx_cluster_sessions_count 1234

    NOTE: Please consider this metric as an approximate estimation. Due to the asynchronous nature of data collection and calculation, exact precision may vary.

• #12338 Introduced a time-based garbage collection mechanism to the RocksDB-based persistent session backend. This feature ensures more efficient management of stored messages, optimizing storage utilization and system performance by automatically purging outdated messages.

• #12398 Exposed the swagger_support option in the Dashboard configuration, allowing for the enabling or disabling of the Swagger API documentation.

• #12467 Started supporting cluster discovery using AAAA DNS record type.

• #12483 Renamed emqx ctl conf cluster_sync tnxid ID to emqx ctl conf cluster_sync inspect ID.

  For backward compatibility, tnxid is kept, but considered deprecated and will be removed in 5.7.

• #12499 Enhanced client banning capabilities with extended rules, including:

  • Matching clientid against a specified regular expression.
  • Matching client's username against a specified regular expression.
  • Matching client's peer address against a CIDR range.

  Important Notice: Implementing a large number of broad matching rules (not specific to an individual clientid, username, or host) may affect system performance. It's advised to use these extended ban rules judiciously to maintain optimal system efficiency.

• #12509 Implemented API to re-order all authenticators / authorization sources.

• #12517 Configuration files have been upgraded to accommodate multi-line string values, preserving indentation for enhanced readability and maintainability. This improvement utilizes """~ and ~""" markers to quote indented lines, offering a structured and clear way to define complex configurations. For example:

  rule_xlu4 {
    sql = """~
      SELECT
        *
      FROM
        "t/#"
    ~"""
  }

  See HOCON 0.42.0 release notes for details.

• #12520 Implemented log throttling. The feature reduces the volume of logged events that could potentially flood the system by dropping all but the first occurance of an event within a configured time window. Log throttling is applied to the following log events that are critical yet prone to repetition:

  • authentication_failure
  • authorization_permission_denied
  • cannot_publish_to_topic_due_to_not_authorized
  • cannot_publish_to_topic_due_to_quota_exceeded
  • connection_rejected_due_to_license_limit_reached
  • dropped_msg_due_to_mqueue_is_full

• #12561 Implemented HTTP APIs to get the list of client's in-flight and message queue (mqueue) messages. These APIs facilitate detailed insights and effective control over message queues and in-flight messaging, ensuring efficient message handling and monitoring.

  To get the first chunk of data:

  • GET /clients/{clientid}/mqueue_messages?limit=100
  • GET /clients/{clientid}/inflight_messages?limit=100

  Alternatively, for the first chunks without specifying a start position:

  • GET /clients/{clientid}/mqueue_messages?limit=100&position=none
  • GET /clients/{clientid}/inflight_messages?limit=100&position=none

  To get the next chunk of data:

  • GET /clients/{clientid}/mqueue_messages?limit=100&position={position}
  • GET /clients/{clientid}/inflight_messages?limit=100&position={position}

  Where {position} is a value (opaque string token) of meta.position field from the previous response.

  Ordering and Prioritization:

  • Mqueue Messages: These are prioritized and sequenced based on their queue order (FIFO), from higher to lower priority. By default, mqueue messages carry a uniform priority level of 0.
  • In-Flight Messages: Sequenced by the timestamp of their insertion into the in-flight storage, from oldest to newest.

• #12590 Removed mfa meta data from log messages to improve clarity.

• #12641 Improved text log formatter fields order. The new fields order is as follows:

  tag > clientid > msg > peername > username > topic > [other fields]

• #12670 Added field shared_subscriptions to endpoint /monitor_current and /monitor_current/nodes/:node.

• #12679 Upgraded docker image base from Debian 11 to Debian 12.

• #12700 Started supporting "b" and "B" unit in bytesize hocon fields. For example, all three fields below will have the value of 1024 bytes:

  bytesize_field = "1024b"
  bytesize_field2 = "1024B"
  bytesize_field2 = 1024

• #12719 The /clients API has been upgraded to accommodate queries for multiple clientids and usernames simultaneously, offering a more flexible and powerful tool for monitoring client connections. Additionally, this update introduces the capability to customize which client information fields are included in the API response, optimizing for specific monitoring needs.

  Examples of Multi-Client/Username Queries:

  • To query multiple clients by ID: /clients?clientid=client1&clientid=client2
  • To query multiple users: /clients?username=user11&username=user2
  • To combine multiple client IDs and usernames in one query: /clients?clientid=client1&clientid=client2&username=user1&username=user2

  Examples of Selecting Fields for the Response:

  • To include all fields in the response: /clients?fields=all (Note: Omitting the fields parameter defaults to returning all fields.)
  • To specify only certain fields: /clients?fields=clientid,username

• #12381 Added new SQL functions: map_keys(), map_values(), map_to_entries(), join_to_string(), join_to_string(), join_to_sql_values_string(), is_null_var(), is_not_null_var().

  For more information on the functions and their usage, refer to Built-in SQL Functions the documentation.

• #12336 Performance enhancement. Created a dedicated async task handler pool to handle client session cleanup tasks.

• #12725 Implemented REST API to list the available source types.

• #12746 Added username log field. If MQTT client is connected with a non-empty username the logs and traces will include username field.

• #12785 Added timestamp_format configuration option to log handlers. This new option allows for the following settings:

  • auto: Automatically determines the timestamp format based on the log formatter being used. Utilizes rfc3339 format for text formatters, and epoch format for JSON formatters.

  • epoch: Represents timestamps in microseconds precision Unix epoch format.

  • rfc3339: Uses RFC3339 compliant format for date-time strings. For example, 2024-03-26T11:52:19.777087+00:00.

Bug Fixes

• #11868 Fixed a bug where will messages were not published after session takeover.

• #12347 Implemented an update to ensure that messages processed by the Rule SQL for the MQTT egress data bridge are always rendered as valid, even in scenarios where the data is incomplete or lacks certain placeholders defined in the bridge configuration. This adjustment prevents messages from being incorrectly deemed invalid and subsequently discarded by the MQTT egress data bridge, as was the case previously.

  When variables in payload and topic templates are undefined, they are now rendered as empty strings instead of the literal undefined string.

• #12472 Fixed an issue where certain read operations on /api/v5/actions/ and /api/v5/sources/ endpoints might result in a 500 error code during the process of rolling upgrades.

• #12492 EMQX now returns the Receive-Maximum property in the CONNACK message for MQTT v5 clients, aligning with protocol expectations. This implementation considers the minimum value of the client's Receive-Maximum setting and the server's max_inflight configuration as the limit for the number of inflight (unacknowledged) messages permitted. Previously, the determined value was not sent back to the client in the CONNACK message.

• #12500 The GET /clients and GET /client/:clientid HTTP APIs have been updated to include disconnected persistent sessions in their responses.

  NOTE: A current known issue with these enhanced API responses is that the total client count provided may exceed the actual number of clients due to the inclusion of disconnected sessions.

• #12513 Changed the level of several flooding log events from warning to info.

• #12530 Improved the error reporting for frame_too_large events and malformed CONNECT packet parsing failures. These updates now provide additional information, aiding in the troubleshooting process.

• #12541 Introduced a new configuration validation step for autocluster by DNS records to ensure compatibility between node.name and cluster.discover_strategy. Specifically, when utilizing the dns strategy with either a or aaaa record types, it is mandatory for all nodes to use a (static) IP address as the host name.

• #12562 Added a new configuration root: durable_storage. This configuration tree contains the settings related to the new persistent session feature.

• #12566 Enhanced the bootstrap file for REST API keys:

  • Empty lines within the file are now skipped, eliminating the previous behavior of generating an error.

  • API keys specified in the bootstrap file are assigned the highest precedence. In cases where a new key from the bootstrap file conflicts with an existing key, the older key will be automatically removed to ensure that the bootstrap keys take effect without issue.

• #12646 Fixed an issue with the rule engine's date-time string parser. Previously, time zone adjustments were only effective for date-time strings specified with second-level precision.

• #12652 Fixed a discrepancy where the subbits functions with 4 and 5 parameters, despite being documented, were missing from the actual implementation. These functions have now been added.

• #12663 Fixed an issue where the emqx_vm_cpu_use and emqx_vm_cpu_idle metrics, accessible via the Prometheus endpoint /prometheus/stats, were inaccurately reflecting the average CPU usage since the operating system boot. This fix ensures that these metrics now accurately represent the current CPU usage and idle, providing more relevant and timely data for monitoring purposes.

• #12668 Refactored the SQL function date_to_unix_ts() by using calendar:datetime_to_gregorian_seconds/1. This change also added validation for the input date format.

• #12672 Changed the process for generating the node boot configuration by incorporating the loading of {data_dir}/configs/cluster.hocon. Previously, changes to logging configurations made via the Dashboard and saved in {data_dir}/configs/cluster.hocon were only applied after the initial boot configuration was generated using etc/emqx.conf, leading to potential loss of some log segment files due to late reconfiguration.

  Now, both {data_dir}/configs/cluster.hocon and etc/emqx.conf are loaded concurrently, with settings from emqx.conf taking precedence, to create the boot configuration.

• #12696 Fixed an issue where attempting to reconnect an action or source could lead to wrong error messages being returned in the HTTP API.

• #12714 Fixed inaccuracies in several metrics reported by the /prometheus/stats endpoint of the Prometheus API. The correction applies to the following metrics:

  • emqx_cluster_sessions_count
  • emqx_cluster_sessions_max
  • emqx_cluster_nodes_running
  • emqx_cluster_nodes_stopped
  • emqx_subscriptions_shared_count
  • emqx_subscriptions_shared_max

  Additionally, this fix rectified an issue within the /stats endpoint concerning the subscriptions.shared.count and subscriptions.shared.max fields. Previously, these values failed to update promptly following a client's disconnection or unsubscription from a Shared-Subscription.

• #12715 Fixed a crash that could occur during configuration updates if the connector for the ingress data integration source had active channels.

• #12740 Fixed an issue when durable sessions could not be kicked out.

• #12768 Addressed a startup failure issue in EMQX version 5.4.0 and later, particularly noted during rolling upgrades from versions before 5.4.0. The issue was related to the initialization of the routing schema when both v1 and v2 routing tables were empty.

  The node now attempts to retrieve the routing schema version in use across the cluster instead of using the v2 routing table by default when local routing tables are found empty at startup. This approach mitigates potential conflicts and reduces the chances of diverging routing storage schemas among cluster nodes, especially in a mixed-version cluster scenario.

  If conflict is detected in a running cluster, EMQX writes instructions on how to manually resolve it in the log as part of the error message with critical severity. The same error message and instructions will also be written on standard error to make sure this message will not get lost even if no log handler is configured.

• #12786 Added a strict check that prevents replicant nodes from connecting to core nodes running with a different version of EMQX application. This check ensures that during the rolling upgrades, the replicant nodes can only work when at least one core node is running the same EMQX release version.

5.5.1

Release Date: 2024-03-06

Bug Fixes

• #12471 Fixed an issue that data integration configurations failed to load correctly during upgrades from EMQX version 5.0.2 to newer releases.

• #12598 Fixed an issue that users were unable to subscribe to or unsubscribe from shared topic filters via HTTP API.

  The affected APIs include:

  • /clients/:clientid/subscribe

  • /clients/:clientid/subscribe/bulk

  • /clients/:clientid/unsubscribe

  • /clients/:clientid/unsubscribe/bulk

• #12601 Fixed an issue where logs of the LDAP driver were not being captured. Now, all logs are recorded at the info level.

• #12606 The Prometheus API experienced crashes when the specified SSL certificate file did not exist in the given path. Now, when an SSL certificate file is missing, the emqx_cert_expiry_at metric will report a value of 0, indicating the non-existence of the certificate.

• #12620 Redacted sensitive information in HTTP headers to exclude authentication and authorization credentials from debug level logs in the HTTP Server connector, mitigating potential security risks.

• #12632 Fixed an issue where the rule engine's SQL built-in function date_to_unix_ts produced incorrect timestamp results for dates starting from March 1st on leap years.

5.5.0

Release Date: 2024-02-01

Enhancements

• #12085 EMQX has been upgraded to leverage the capabilities of OTP version 26.1.2-2. NOTE: Docker images are still built with OTP 25.3.2.

• #12189 Enhanced the ACL claim format in EMQX JWT authentication for greater versatility. The updated format now supports an array structure, aligning more closely with the file-based ACL rules.

  For example:

  [
  {
    "permission": "allow",
    "action": "pub",
    "topic": "${username}/#",
    "qos": [0, 1],
    "retain": true
  },
  {
    "permission": "allow",
    "action": "sub",
    "topic": "eq ${username}/#",
    "qos": [0, 1]
  },
  {
    "permission": "deny",
    "action": "all",
    "topics": ["#"]
  }
  ]

  In this new format, the absence of a matching rule does not result in an automatic denial of the action. The authorization chain can advance to other configured authorizers if a match is not found in the JWT ACL. If no match is found throughout the chain, the final decision defers to the default permission set in authorization.no_match.

• #12267 Added a new timeout parameter to the cluster/:node/invite interface, addressing the issue of default timeouts. The previously set 5-second default timeout often led to HTTP API call timeouts because joining an EMQX cluster usually requires more time.

  In addition, EMQX added a new API /cluster/:node/invite_async to support an asynchronous way to invite nodes to join the cluster and introduced a new cluster/invitation API to inspect the join status.

• #12272 Introduced updates to the retain API in EMQX:

  • Added a new API DELETE /retainer/messages to clean all retained messages.
  • Added an optional topic filter parameter topic in the query string for the API GET /retainer/messages. For example, using a query string topic=t/1 filters the retained messages for a specific topic, improving the efficiency of message retrieval.

• #12277 Added mqtt/delayed/messages/:topic API to remove delayed messages by topic name.

• #12278 Adjusted the maximum pagination size for paginated APIs in the REST API from 3000 to 10000.

• #12289 Authorization caching now supports the exclusion of specific topics. For the specified list of topics and topic filters, EMQX will not generate an authorization cache. The list can be set through the authorization.cache.excludes configuration item or via the Dashboard. For these specific topics, permission checks will always be conducted in real-time rather than relying on previous cache results, thus ensuring the timeliness of authorization outcomes.

• #12329 Added broker.routing.batch_sync configuration item to enable a dedicated process pool that synchronizes subscriptions with the global routing table in batches, thus reducing the frequency of cross-node communication that can be slowed down by network latency. Processing multiple subscription updates collectively, not only accelerates synchronization between replica nodes and core nodes in a cluster but also reduces the load on the broker pool, minimizing the risk of overloading.

• #12333 Added a tags field for actions and connectors. Similar to the description field (which is a free text annotation), tags can be used to annotate actions and connectors for filtering and grouping.

• #12299 Exposed more metrics to improve observability:

  Montior API:

  • Added retained_msg_count field to /api/v5/monitor_current.
  • Added license_quota field to /api/v5/monitor_current
  • Added retained_msg_count and node_uptime fields to /api/v5/monitor_current/nodes/{node}.
  • Added retained_msg_count, license_quota and node_uptime fields to /api/v5/monitor_current/nodes/{node}.

  Prometheus API:

  • Added emqx_cert_expiry_at and emqx_license_expiry_at to /api/v5/prometheus/stats to display TLS listener certificate expiration time and license expiration time.
  • Added /api/v5/prometheus/auth endpoint to provide metrics such as execution count and running status for all authenticatiors and authorizators.
  • Added /api/v5/prometheus/data_integration endpoint to provide metrics such as execution count and status for all rules, actions, and connectors.

  Limitations: Prometheus push gateway only supports the content in /api/v5/prometheus/stats?mode=node.

  For more API details and metric type information, please see swagger api docs.

• #12196 Improved network efficiency during routes cleanup.

  Previously, when a node was down, a delete operation for each route to that node must be exchanged between all the other live nodes. After this change, only one match and delete operation is exchanged between all live nodes, significantly reducing the number of necessary network packets and decreasing the load on the inter-cluster network. This optimization must be especially helpful for geo-distributed EMQX deployments where network latency can be significantly high.

• #12354 The concurrent creation and updates of data integrations are now supported, significantly increasing operation speeds, such as when importing backup files.

Bug Fixes

• #12232 Fixed an issue when cluster commit log table was not deleted after a node was forced to leave a cluster.

• #12243 Fixed a family of subtle race conditions that could lead to inconsistencies in the global routing state.

• #12269 Improved error handling in the /clients interface; now returns a 400 status with more detailed error messages, instead of a generic 500, for query string validation failures.

• #12285 Updated the CoAP gateway to support short parameter names for slight savings in datagram size. For example, clientid=bar can be written as c=bar.

• #12303 Fixed the message indexing in retainer. Previously, clients with wildcard subscriptions might receive irrelevant retained messages not matching their subscription topics.

• #12305 Corrected an issue with incomplete client/connection information being passed into emqx_cm, which could lead to internal inconsistencies and affect memory usage and operations like node evacuation.

• #12306 Fixed an issue preventing the connectivity test for the Connector from functioning correctly after updating the password parameter via the HTTP API.

• #12359 Fixed an issue causing error messages when restarting a node configured with some types of data bridges. Additionally, these bridges were at risk of entering a failed state upon node restart, requiring a manual restart to restore functionality.

• #12404 Fixed an issue where restarting a data integration with heavy message flow could lead to a stop in the collection of data integration metrics.

5.4.1

Release Date: 2024-01-09

Bug Fixes

• #12234 Resolved compatibility issues with Open Telemetry configurations defined in emqx.conf from versions before EMQX 5.4.0, ensuring smooth integration of legacy configurations with the latest EMQX release.
• #12236 Fixed client ID generation in MQTT broker data integration to comply with MQTT 3.1 specification of 23-byte limit. Client ID is now prefixed with user-assigned Connector name, followed by the first 8 bytes of node name's SHA hash and pool member ID. If the resulting ID exceeds 23 bytes, additional SHA hash and truncation for the first 23 characters are applied to ensure compliance.
• #12238 Resolved compatibility issue with the error format configurations introduced in the HTTP Action feature of EMQX version 5.3.2.
• #12246 Stopped exposing port 11883 by default in Docker and removed it from Helm charts, as this port is no longer in use.
• #12249 Fixed an issue in the /configs API where attempting to modify a read-only configuration value resulted in a garbled response message.
• #12264 Fixed an issue where version 5.4 replica nodes could not join clusters with core nodes running versions earlier than 5.4 during the rolling upgrade process.

5.4.0

Release Date: 2023-12-23

Enhancements

• #11884 Modified the Prometheus API and configuration to implement the following improvements:

  • Restructured configuration sections to group-related settings, improving readability and maintainability.
  • Introduced enable_basic_auth configuration for basic authentication on the scrape API endpoint, enhancing security.
  • Maintained backward compatibility while refactoring code, avoiding breaking changes.

• #11896 Introduced an enhancement for configuring sensitive authentication fields in bridges, such as passwords, tokens, and secret keys. This improvement allows the use of secrets stored as files in the file system. These secrets can be securely referenced in configuration files using the special file:// prefix, enhancing the security of sensitive data handling in bridge configurations.

• #11921 Introduced Open Telemetry Logs Handler that allows to format log events in alignment with the Open Telemetry log data model. This handler facilitates the exportation of formatted log events to a configured Open Telemetry collector or back-end, thereby enhancing log management and integration capabilities.

• #11935 Switched to the new v2 routing store schema by default. The new schema improves both subscription and routing performance, especially in scenarios with concurrent subscriptions to topic filters sharing common wildcard prefixes. However, it does come with a minor increase in memory usage. This schema also eliminates the need for a separate index, thus inconsistencies in the routing state rarely encountered in previous versions should no longer be possible.

  If a cluster is rolling upgraded from an older version, the cluster will continue to use v1 store until a full cluster (non-rolling) restart happens.

  Users can still opt for the previous schema by configuring the broker.routing.storage_schema option to v1. However, this also requires a complete, non-rolling restart of the cluster to take effect.

• #11984 Implemented Open Telemetry distributed tracing feature.

• #12017 Implemented a dedicated REST API for the import and export of configuration and user data.

• #12040 Upgraded QUIC protocol stack.

• #12201 Added support for hot updates to TCP/SSL/WS/WSS MQTT listener configurations. This feature allows you to modify most configuration parameters without restarting the listener and disconnecting the clients. However, there are some limitations:

  • For TCP/SSL listeners, changes to the following parameters will still require a listener restart and client reconnection:

    • bind
    • tcp_options.backlog

  • For WS/WSS (WebSocket) listeners, modifying transport-related parameters (listed below) will result in the listening socket being reopened, but established connections will remain uninterrupted.

    • bind

    • tcp_options.*

    • ssl_options.*

• #11608 Integrated LDAP bind operation as a new authenticator, providing a more flexible and secure method for user authentication.

• #11766 Implemented a preliminary Role-Based Access Control for the REST API. In this version, there are three predefined roles:

  • Administrator: This role can access all resources.

  • Viewer: This role can only view resources and data, corresponding to all GET requests in the REST API.

  • Publisher: Specifically tailored for MQTT message publishing, this role is confined to accessing endpoints related to message publication.

• #11773 Implemented Dashboard support for audit log management. Users can utilize this page to view all change operations performed on EMQX devices and data, such as kicking out devices, creating/deleting rules, etc.

• #11778 Integrated Microsoft Entra Identity (formerly known as Azure Active Directory) support into the SAML single sign-on (SSO) process.

• #11811 Improved the format for the REST API key bootstrap file to support initializing key with a role.

  The new form is:api_key:api_secret:role.

  role is optional and its default value is administrator.

• #11852 Introduced a new GB/T 32960 gateway, enabling vehicles to connect with EMQX via the GBT32960 vehicular networking protocol.

• #11883 Introduced a new JT/T808 gateway, enabling vehicles to connect with EMQX via the JT/T 808 vehicular networking protocol.

• #11885 Introduced a new OCPP gateway for Electric vehicle (EV) charging stations to access EMQX through the OCPP (Open Charge Point Protocol).

• #11971 Made /api/v5/load_rebalance/availability_check public, meaning it no longer requires authentication. This change simplifies the setup of load balancers.

  It improved the gracefulness of the rebalance/evacuation process during the wait health check phase. The connections to nodes marked for eviction are now not prohibited during this phase. During this phase it is unknown whether these nodes are all marked unhealthy by the load balancer, so prohibiting connections to them may cause multiple unsuccessful reconnection attempts.

• #12013 The data bridging design has been adjusted to split it into connectors and actions (Sinks). Connectors are used to manage the integration of data with external systems and can be reused across multiple actions, while actions are used to configure how data is processed. This design provides greater flexibility and scalability, resulting in clearer data integration configuration and management.

  The adjusted data bridges includes PostgreSQL, Timescale, and Matrix, which have now been split into connectors and actions APIs, but they remain backward compatible with the old data bridge API.

• #12016 Enhanced license key management.

  EMQX can now load the license key from a specified file. This is enabled by setting the license.key configuration to a file path, which should be prefixed with "file://". Also added the ability to revert to the default trial license by setting license.key = default. This option simplifies the process of returning to the trial license if needed.

• #12129 Renewed the default license, replacing the old license issued in January 2023. At the same time, the license capacity has been adjusted from 100 concurrent connections to 25 concurrent connections.

Bug Fixes

• #10976 Fixed topic-filter overlapping handling in shared subscription. In the previous implementation, the storage method for subscription options did not provide adequate support for shared subscriptions. This resulted in message routing failures and leakage of routing tables between nodes during the "subscribe-unsubscribe" process with specific order and topics.

• #12048 Fixed COAP gateway bug that caused it to ignore subscription options.

• #12078 Upgraded grpc-erl to 0.6.12. This update addresses a potential deadlock issue where the grpc client started dependent apps lazily.

• #12081 Updated gen_rpc library to version 3.3.1. The new version includes several performance improvements:

  • Avoiding allocating extra memory for the packets before they are sent to the wire in some cases.

  • Bypassing network for the local calls.

  • Avoid senstive data leaking in debug logs #12202

• #12111 Fixed an issue when API tokens were sometimes unavailable immediately after login due to race condition.

• #12121 Fixed an issue where nodes in the cluster would occasionally return a stale view when updating configurations on different nodes concurrently.

• #12158 Fixed an issue when the rule engine cannot connect to Redis hosted by Upstash.

  Before the fix, after establishing a TCP connection with the Redis service, the Redis driver of EMQX used Inline Commands to send AUTH and SELECT commands. However, the upstash Redis service does not support Inline Commands, which causes the rule engine to fail to connect to the upstash Redis service. After the fix, the Redis driver of EMQX uses RESP (REdis Serialization Protocol) to send AUTH and SELECT commands.

• #12176 Always acknowledge DISCONNECT packet to MQTT-SN client regardless of whether the connection has been successfully established before.

• #12180 Fix an issue where DTLS enabled MQTT-SN gateways could not be started, caused by incompatibility of default listener configuration with the DTLS implementation.

• #12219 Fix file transfer S3 config secret deobfuscation issue while performing config updates from dashboard.

5.3.2

Release Date: 2023-12-01

Enhancements

• #11752 Changed default RPC driver from gen_rpc to rpc for core-replica database synchronization.

  This improves core-replica data replication latency.

• #11785 Allowed users with the "Viewer" role to change their own passwords. However, those with the "Viewer" role do not have permission to change the passwords of other users.

• #11787 Improved the performance of the emqx command.

• #11790 Added validation to Redis commands in Redis authorization source. Additionally, this improvement refines the parsing of Redis commands during authentication and authorization processes. The parsing now aligns with redis-cli compatibility standards and supports quoted arguments.

• #11541 Enhanced file transfer capabilities. Now, clients can use an asynchronous method for file transfer by sending commands to the $file-async/... topic and subscribing to command execution results on the $file-response/{clientId} topic. This improvement simplifies the use of the file transfer feature, particularly suitable for clients using MQTT v3.1/v3.1.1 or those employing MQTT bridging. For more details, please refer to EIP-0021.

Bug Fixes

• #11757 Fixed the error response code when downloading non-existent trace files. Now the response returns 404 instead of 500.

• #11762 Fixed an issue in EMQX's built_in_database authorization source. With this update, all Access Control List (ACL) records are completely removed when an authorization source is deleted. This resolves the issue of residual records remaining in the database when re-creating authorization sources.

• #11771 Fixed validation of Bcrypt salt rounds in authentication management through the API/Dashboard.

• #11780 Fixed validation of the iterations field of the pbkdf2 password hashing algorithm. Now, iterations must be strictly positive. Previously, it could be set to 0, which led to a nonfunctional authenticator.

• #11791 Fixed an issue in the EMQX CoAP Gateway where heartbeats were not effectively maintaining the connection's active status. This fix ensures that the heartbeat mechanism properly sustains the liveliness of CoAP Gateway connections.

• #11797 Modified HTTP API behavior for APIs managing the built_in_database authorization source. They will now return a 404 status code if built_in_database is not set as the authorization source, replacing the former 20X response.

• #11965 Improved the termination of EMQX services to ensure a graceful stop even in the presence of an unavailable MongoDB resource.

• #11975 This fix addresses an issue where redundant error logs were generated due to a race condition during simultaneous socket closure by a peer and the server. Previously, concurrent socket close events triggered by the operating system and EMQX resulted in unnecessary error logging. The implemented fix improves event handling to eliminate unnecessary error messages.

• #11987 Fixed a bug where attempting to set the active_n option on a TCP/SSL socket could lead to a connection crash.

  The problem occurred if the socket had already been closed by the time the connection process attempted to apply the active_n setting, resulting in a case_clause crash.

• #11731 Added hot configuration support for the file transfer feature.

• #11754 Improved the log formatting specifically for the Postgres bridge in EMQX. It addresses issues related to Unicode characters in error messages returned by the driver.

5.3.1

Release Date: 2023-11-14

Enhancements

• #11637 Added extra diagnostic checks to help debug issues when mnesia is stuck waiting for tables. Library Updates: ekka has been upgraded to version 0.15.15, and mria to version 0.6.4.
• #11581 Feature Preview: Planned for EMQX v5.4.0, introducing the concepts of Connector and Action base on data bridge. The existing data bridge will be gradually migrated to Connector and Action. Connector are designed to manage the integration with external systems, while Actions are solely used to configure the data processing methods. Connector can be reused across multiple Actions, providing greater flexibility and scalability. Currently, the migration has been completed for Kafka producer and Azure Event Hub producer.
• The Dashboard now supports MQTT 5.0 publish attribute settings for the rule engine's message republish action, allowing users more flexibility in publishing messages.

Bug Fixes

• #11565 Upgraded jq library from v0.3.10 to v0.3.11. In this version, jq_port programs are initiated on-demand and will not appear in users' processes unless the jq function in EMQX is used. Additionally, idle jq_port programs will auto-terminate after a set period. Note: Most EMQX users are running jq in NIF mode and will not be affected by this update.

• #11676 Hid a few pieces of sensitive information from debug-level logs.

• #11697 Disabled outdated TLS versions and cipher suites in the EMQX backplane network (gen_rpc). Added support for tlsv1.3 on the backplane and introduced new configuration parameters: EMQX_RPC__TLS_VERSIONS and EMQX_RPC__CIPHERS.

  The corresponding gen_rpc PR: https://github.com/emqx/gen_rpc/pull/36

• #11734 Fixed clustering in IPv6 network. Added new configurations rpc.listen_address and rpc.ipv6_only to allow EMQX cluster RPC server and client to use IPv6.

• #11747 Updated QUIC stack to msquic 2.2.3.

• #11796 Fixed rpc schema to ensure that client/server uses same transport driver.

• #11798 Fixed the issue where the node could not start after executing ./bin/emqx data import [FILE].

  The connection between apikey_key and apikey_name is also enhanced for better consistency and unique identification.

  • apikey_key: When generating an API key via the dashboard, apikey_key will now create a unique value derived from the provided human-readable apikey_name.
  • apikey_name Conversely, when using a bootstrap file to generate an API key, apikey_name will be generated as a unique value based on the associated apikey_key.

• #11813 Fixed the schema to ensure that RPC client SSL port aligns with the configured server port. This fix also guarantees that the RPC ports are correctly opened in the Helm chart.

• #11819 Upgraded opentelemetry library to v1.3.1-emqx. This opentelemetry release fixes invalid metrics timestamps in the exported metrics.

• #11861 Fixed excessive warning message printed in remote console shell.

• #11722 Fixed an issue where a Kafka Producer bridge with sync query mode would not buffer messages when in the connecting state.

• #11724 Fixed a metrics-related issue where messages sent to Kafka would be counted as failed even when they were successfully transmitted afterward due to internal buffering.

• #11728 Enhanced the LDAP filter string parser with the following improvements:

  • Automatic escaping of special characters within filter strings.
  • Fixed a bug that previously prevented the use of dn as a filter value.

• #11733 Resolved an incompatibility issue that caused crashes during session takeover or channel eviction when the session was located on a remote node running EMQX v5.2.x or an earlier version.

• #11750 Eliminated logging and tracing of HTTP request bodies in HTTP authentification and HTTP bridges.

• #11760 Simplified the CQL query used for the Cassandra bridge health check, which was previously generating warnings in the Cassandra server logs.

• #11886 Fixed backward plugin compatibility.

  Currently, EMQX validates hook point names, and invalid hook points cannot be used for hook registration. However, some older versions of plugin templates used misspelled hook points, and actual plugins in use may also have this issue. To maintain compatibility with these older plugins, we allow the use of the old hook points for hook registration, but we issue deprecated warnings for them. As before, these hooks will not be called.

• #11897 Fixed the issue of waiting for a loop race condition during node configuration synchronization when cluster nodes are started approximately at the same time.

5.3.0

Release Date: 2023-09-29

Enhancements

• #11597 Upgraded ekka to 0.15.13, which incorporates the following changes:

  • Upgraded Mria to 0.6.2.
  • Introduced the ability to configure the bootstrap data sync batch size, as detailed in Mria PR.
  • Enhanced the reliability of mria_membership processes, as described in Mria PR.
  • Fix log message formatting error.
  • Added node.default_bootstrap_batch_size option to EMQX configuration. Increasing the value of this option can greatly reduce a replicant node startup time, especially when the EMQX cluster interconnect network latency is high and the EMQX built-in database holds a large amount of data, e.g. when the number of subscriptions is high.

• #11620 Added a new rule-engine SQL function bytesize to get the size of a byte-string. e.g. SELECT * FROM "t/#" WHERE bytesize(payload) > 10.

• #11642 Updated to quicer version 0.0.200 in preparation for enabling openssl3 support for QUIC transport.

• #11610 Implemented a preliminary Role-Based Access Control for the Dashboard.

  In this version, there are two predefined roles:

  • Administrator: This role could access all resources.

  • Viewer: This role can only view resources and data, corresponding to all GET requests in the REST API.

• #11631 Added Single Sign-On (SSO) feature and integrated with LDAP.

• #11656 Integrated the SAML 2.0 Support for SSO.

• #11599 Supported audit logs to record operations from CLI, REST API, and Dashboard in separate log files.

Bug Fixes

• #11682 Fixed an issue where logging would stop if "Rotation Size" would be set to infinity on file log handlers.
• #11567 Improve EMQX graceful shutdown (emqx stop command):
  • Increase timeout from 1 to 2 minutes.
  • Printed an error message if EMQX can't stop gracefully within the configured timeout.
  • Print periodic status messages while EMQX is shutting down.
• #11584 Fixed telemetry reporting error on Windows when os_mon module is unavailable.
• #11605 Lowered CMD_overridden log severity from warning to info.
• #11622 Upgraded rpc library gen_rpc from 2.8.1 to 3.1.0.
• #11623 Upgraded library esockd from 5.9.6 to 5.9.7. This upgrade included:
  • Enhancements regarding proxy protocol error and timeout. esockd pr#178
  • Lowered ssl_error exceptions to info-level logging. esockd pr#180
  • Malformed MQTT packet parsing exception log level is lowered from error to info.
  • In command emqx ctl listeners output, the shutdown_count counter is incremented when TLS handshake failure (ssl_error) or Malformed packet (frame_error) happens.
• #11661 Fixed log formatter when log.HANDLER.formatter is set to 'json'. The bug was introduced in v5.0.4 where the log line was no longer a valid JSON, but prefixed with timestamp string and level name.
• #11627 Fixed resources cleanup in HStreamdB bridge. Prior to this fix, HStreamDB bridge might report errors during bridge configuration updates, since hstreamdb client/producer were not stopped properly.

5.2.1

Release Date: 2023-09-20

Enhancements

• #11487 The bcrypt work factor is limited to the range 5-10, because higher values consume too much CPU resources. Bcrypt library is updated to allow parallel hash evaluation.

• #11568 Added support for defining templates for MQTT 5.0 publish properties and user properties in Republish rule action.

• #11612 During node evacuation, evacuate all disconnected sessions, not only those started with clean_start set to false.

• #11532 Improved error messaging for better clarity when parsing invalid packets.

Bug Fixes

• #11493 Fixed response examples for /api/v5/publish bad request in RESP API documentation. Previously the documentation example said that the bad request response could return a list in the body which was not actually the case.

• #11499 Upgraded Erlang/OTP to version 25.3.2-2, which now excludes sensitive data from mnesia_hook log messages.

• #11506 Previously, attempting to download a non-existent trace log file would result in downloading an empty file. After implementing this fix, when attempting to download an empty trace log file using the GET request /api/v5/trace/clientempty/download, the server will now respond with a 404 status code and the following JSON message: {"code":"NOT_FOUND","message":"Trace is empty"}. This response will be triggered if no events matching the trace condition are found in the log file.

• #11522 Improved rule engine schema registry error message when schema name exceeds the permissible length.

• #11531 Fixed an issue where authorization cache cleaning CLI was not working properly for specific client ID.

• #11564 Fixed cluster partition autoheal functionality. Implemented autohealing for the clusters that split into multiple partitions.

• #11568 Fixed an issue where an ill-defined built-in rule action config could be interpreted as a custom user function.

• #11394 Upgraded Kafka producer client wolff from 1.7.6 to 1.7.7. This fixed a potential race condition that might cause all Kafka producers to crash if some failed to initialize.

• #11401 Fixed the behavior of the rule SQL mongo_date function in SQL statement testing in the EMQX Dashboard. The rule SQL mongo_date function now returns a string with the format ISODate(*), where * is an ISO date string when running rules in test mode. This format aligns with how MongoDB stores dates.

• #11547 Fixed several emqx_bridge issues:

  • Fixed Cassandra bridge connect error occurring when the bridge is configured without username/password (Cassandra doesn't require user credentials when it is configured with authenticator: AllowAllAuthenticator.)
  • Fixed SQL Server bridge connect error caused by an empty password.
  • Made username a required field in Oracle bridge.
  • Fixed IoTDB bridge error caused by setting base URL without a scheme (e.g. <host>:<port>).

• #11630 Fixed an issue where the core node could get stuck in the mria_schema:bootstrap/0 state, preventing new nodes from joining the cluster.

5.2.0

Release Date: 2023-09-07

Enhancements

• #10697 This enhancement enables the configuration of the minReadySeconds for the StatefulSet. This feature allows for the introduction of a time gap between the restarts of individual pods triggered by upgrade or restart commands.

• #11124 Released packages for Amazon Linux 2023.

• #11289 Released packages for Debian 12.

• #11290 Updated the jq dependency to version 0.3.10, which includes an update to the oniguruma library to version 6.9.8 with a few minor security fixes.

• #11291 Updated RocksDB version to 1.8.0-emqx-1 via ekka update to 0.15.6.

• #11390 Added node.broker_pool_size, node.generic_pool_size, node.channel_cleanup_batch_size options to EMQX configuration. Tuning these options can significantly improve the performance if cluster interconnect network latency is high.

• #11429 Added an option to configure detection of the legacy protocol in MondoDB connectors and bridges.

• #11436 Added a new API endpoint DELETE/banned for clearing all banned data.

• #11438 Changed the type of the mqtt.max_packet_size from string to byteSize for a better representation of the valid numeric range. Strings will still be accepted for backward compatibility.

• #11469 Added support for specifying username in Redis authentication.

• #11496 Disabled the Erlang VM Prometheus exporter by default to improve performance and security.

• #11497 Enhanced broker metrics collection and export by adding new metrics for messages, overload protection, authorization, authentication, and improving naming consistency for OpenTelemetry.

• #10647 Implemented GreptimeDB data integration.

• #11261 Implemented Amazon Kinesis Data Streams producer data integration.

• #11329 Implemented Azure Event Hub Producer data integration.

• #11363 Added TLS connection support to the RabbitMQ bridge.

• #11367 Ported GCP IoT Hub authentication support from EMQX 4.4.

• #11386 Integrated LDAP as a new authenticator.

• #11392 Integrated LDAP as an authorization source.

• #11402 Added support for using placeholders to define MQTT Topic in Kafka Consumer bridge topic mappings. This allows dynamically setting the MQTT Topic.

• #11403 Added support for defining message attributes and ordering key templates for GCP PubSub Producer bridge.

  Also updated our HOCON library to fix an issue where objects in an array were concatenated even if they were laid on different lines.

• #11459 Added the option to configure health check interval for Kafka bridges.

• #11478 Added HStreamDB bridge support (both TCP and TLS connection allowed), adapted to the HStreamDB v0.16.1.

  Updated driver to 0.4.5+v0.16.1 in PR#11530.

• #11389 Improved retained message publishing latency by consolidating multiple index update operations into a single Mnesia activity, leveraging the new APIs introduced in Mria 0.6.0.

• #11396 Introduced topic index for the rule engine runtime to speed up matching messages' topics to topic filters configured in rule definitions by avoiding full scan of the rule set, significantly improving EMQX's performance when handling a substantial number of rules.

• #11399 Improved the placeholder syntax in the rule engine. The republishing actions support placeholder syntax to dynamically fill in the content of strings in the payload variable. The format of the placeholder syntax is \${key}. Before this improvement, the key in \${key} could only contain letters, numbers, and underscores. Now the key supports any UTF8 characters.

• #11405 Made the error message for date_to_unix_ts function more understandable.

• #11490 Added fast error handling for undefined passwords in various authentication backends. This improves the consistency and user-friendliness of the authentication process.

Bug Fixes

• #11065 Silenced irrelevant error messages during EMQX shutdown.

• #11279 Fixed an issue where clients could not send messages with large payloads when debug/trace logging was enabled in EMQX.

• #11296 Added support for importing additional configurations from EMQX backup file using the emqx ctl import command):

  • rule_engine (previously not imported due to the bug)
  • topic_metrics (previously not implemented)
  • slow_subs (previously not implemented)

• #11327 Updated ekka to version 0.15.8, mria to version 0.15.8, and optvar to 1.0.5. This fixes occasional assertion failures.

• #11346 Updated ekka to version 0.15.9. This fixes dangling etcd locks that occurred when acquiring the lock failed with a timeout.

• #11347 Ensured that OCSP request path is properly URL encoded.

• #11352 Fixed a crash issue that occurred when starting on Windows or any other platform without RocksDB support.

• #11388 Increased emqx_router_sup restart intensity to improve tolerance for occasional crashes that can occur under normal conditions, without necessitating the shutdown of the entire EMQX application. For example, mria write/delete call delegated from a replicant to a core node by emqx_router_helper may fail, if the core node undergoes stopping, restarting, or is in an unready state. The modified restart intensity ensures that the system remains stable and operational.

  This fixes issues found when trying to upgrade from 5.1.3 where that option was set in the configuration files or persisted in EMQX Operator settings.

• #11424 Added a check for the maximum value of the timestamp in the API to ensure it is a valid Unix timestamp.

• #11445 Removed os_mon application monitor support on Windows platforms to prevent VM crashes. Functionality remains on non-Windows platforms.

• #11454 Fixed crashing when debugging/tracing with large payloads (introduced in #11279).

• #11456 Removed validation that enforced non-empty PEM for the CA cert file, allowing the CA certificate file PEM to be empty.

• #11466 Fixed a crash that occurred when setting the ssl_options.ciphers configuration option to an empty string ("").

• #11480 Improves the error handling and testing of SQL functions in the rule engine when rule functions receive bad arguments.

• #11520 Fixed issue where packets_connack_sent metric was not incremented on CONNACK packets sent with non-zero ack_flag.

• #11523 Corrected a misleading prompt when specifying invalid certificates/keys for the /configs API.

• #11534 Fixed the increment on data bridge statistics when the bridge is unhealthy. Now, messages sent to unhealthy bridges are counted as dropped messages.

• #11540 Improved HTTP response when attempting to create a bridge with an invalid name.

• #11548 Fixed an issue that prevented the plugin order from being updated across the entire cluster.

• #11366 Fixed an issue that could prevent a pod from starting if some bridge configurations were specified in bootstrapConfig using EMQX Operator.

• #11453 Fixed an issue that would yield false negatives when testing the connectivity of InfluxDB bridges.

• #11461 Aligned the timeout for testing bridge connectivity more closely with the configured health check timeout.

• #11492 Fixed an issue that would yield false negatives when testing the connectivity of GreptimeDB bridges.

• #11508 Fixed error handling in Kafka bridge when headers are translated to an invalid value.

• #11513 Fixed a bug that prevented the Kafka Producer bridge from using the correct template for the timestamp field.

• #11527 Fixed an issue related to Kafka header template handling. The issue occurs when placeholders are resolved into an array of key-value pairs (e.g.: [{"key": "foo", "value": "bar"}]).

5.1.1

Release Date: 2023-07-05

Enhancements

• #10667 The MongoDB connector and bridge have been refactored into a separate app to improve the code structure.

• #11115 Added info logs to indicate when buffered messages are dropped due to time-to-live (TTL) expiration.

• #11133 Renamed deliver_rate to delivery_rate in the configuration of retainer, while being compatible with the previous deliver_rate.

• #11137 Refactored the Dashboard listener configuration to use a nested ssl_options field for SSL settings.

• #11138 Changed the default value of k8s api_server from http://127.0.0.1:9091 to https://kubernetes.default.svc:443.

  • emqx_ctl conf show cluster no longer displays irrelevant configuration items when discovery_strategy=static. Configuration information related to etcd/k8s/dns will not be shown.
  • Removed zones (deprecated config key) from emqx_ctl conf show_keys.

• #11165 Removed the /configs/limiter API from swagger.json. Only the API documentation was removed, and the /configs/limiter API functionalities remain unchanged.

• #11166 Added 3 random SQL functions to the rule engine:

  • random(): Generates a random number between 0 and 1 (0.0 =< X < 1.0).
  • uuid_v4(): Generates a random UUID (version 4) string.
  • uuid_v4_no_hyphen(): Generates a random UUID (version 4) string without hyphens.

• #11180 Added a new configuration API /configs (GET/PUT) that supports reloading the HOCON format configuration file.

• #11226 Unified the listener switch to enable, while being compatible with the previous enabled.

• #11249 Added /license/setting REST API endpoint to read and update licensed connections usage alarm watermark.

• #11251 Added the /cluster/topology REST API endpoint:

  A GET request to this endpoint returns the cluster topology, showing connections between RLOG core and replicant nodes.

• #11253 The Webhook/HTTP bridge has been refactored into its own Erlang application. This allows for more flexibility in the future and allows the bridge to be run as a standalone application.

• #11079 Added support for custom headers in messages for Kafka bridge producer mode.

• #11132 Added support for MQTT action authorization based on QoS level and Retain flag values. Now, EMQX can verify whether clients have the permission to publish/subscribe using specific QoS levels, and whether they have the permission to publish retained messages.

• #11207 Updated the driver versions of multiple data bridges to enhance security and ensure that sensitive data will not be leaked. This includes:

  • TDengine
  • MongoDB
  • MySQL
  • Clickhouse

• #11241 Schema Registry has been refactored into its own Erlang application. This allows for more flexibility in the future.

• #11020 Upgraded emqtt dependency to prevent sensitive data leakage in the debug log.

• #11135 Improved time offset parser in rule engine and return uniform error codes.

• #11236 Improved the speed of clients querying in REST API /clients endpoint with default parameters.

Bug Fixes

• #11004 Wildcards are no longer allowed for the destination topic in topic rewrite.

• #11026 Addressed an inconsistency in the usage of div and mod operations within the rule engine. Previously, the div' operation could only be used as an infix operation, and mod could only be applied through a function call. Now, both div and mod can be used via function call syntax and infix syntax.

• #11037 When starting an HTTP connector, EMQX now returns a descriptive error in case the system is unable to connect to the remote target system.

• #11039 Fixed database number validation for Redis connector. Previously, negative numbers were accepted as valid database numbers.

• #11074 Fixed a bug to adhere to Protocol spec MQTT-5.0 [MQTT-3.8.3-4].

• #11077 Fixed a crash when updating listener binding with a non-integer port.

• #11094 Fixed an issue where connection errors in Kafka Producer would not be reported when reconnecting the bridge.

• #11103 Updated erlcloud dependency.

• #11106 Added validation for the maximum number of worker_pool_size of a bridge resource.

  Now the maximum amount is 1024 to avoid large memory consumption from an unreasonable number of workers.

• #11118 Ensured that validation errors in REST API responses are slightly less confusing. Now, if there are out-of-range errors, they will be presented as {"value": 42, "reason": {"expected": "1..10"}, ...}, replacing the previous usage of expected_type with expected.

• #11126 Rule metrics for async mode bridges will set failure counters correctly now.

• #11134 Fixed the value of the uppercase authorization header not being obfuscated in the log.

• #11139 The Redis connector has been refactored into its own Erlang application to improve the code structure.

• #11145 Added several fixes and improvements in Ekka and Mria.

  Ekka:

  • Improved cluster discovery log messages to consistently describe actual events. Ekka PR
  • Removed deprecated cluster auto-clean configuration parameter (it has been moved to Mria). Ekka PR

  Mria:

  • Ping now only runs on replicant nodes. Previously, mria_lb was trying to ping both stopped and running replicant nodes, which could result in timeout errors. Mria PR
  • Used null_copies storage when copying $mria_rlog_sync table. This fix has no effect on EMQX for now, as $mria_rlog_sync is only used in mria:sync_transaction/2,3,4, which is not utilized by EMQX. Mria PR

• #11148 Fixed an issue when nodes tried to synchronize configuration update operations to a node which has already left the cluster.

• #11150 Wait for Mria table when emqx_psk app is being started to ensure that PSK data is synced to replicant nodes even if they don't have init PSK file.

• #11151 The MySQL connector has been refactored into its own Erlang application to improve the code structure.

• #11158 Wait for Mria table when the mnesia backend of retainer starts to avoid a possible error of the retainer when joining a cluster.

• #11162 Fixed an issue in webhook bridge where, in async query mode, HTTP status codes like 4XX and 5XX would be treated as successes in the bridge metrics.

• #11164 Reintroduced support for nested (i.e.: ${payload.a.b.c}) placeholders for extracting data from rule action messages without the need for calling json_decode(payload) first.

• #11172 Fixed the payload field in rule engine SQL being duplicated in the below situations:

  • When using a foreach sentence without the as sub-expression and selecting all fields (using the * or omitting the do sub-expression).

  For example:

  FOREACH payload.sensors FROM "t/#"

  • When selecting the payload field and all fields.

  For example:

  SELECT payload.sensors, * FROM "t/#"

• #11174 Fixed the encoding of the server key coming from an ingress MQTT bridge.

  Before the fix, it was encoded as a list of integers corresponding to the ASCII characters of the server string.

• #11184 Config value for mqtt.max_packet_size now has a max value of 256MB as defined by the protocol.

• #11192 Fixed an issue with producing invalid HOCON file when an atom type was used. Also removed unnecessary " around keys and latin1 strings from HOCON file.

• #11195 Fixed an issue where the REST API could create duplicate subscriptions for specified clients of the Stomp gateway.

• #11206 Made the username and password params of CoAP client optional in connection mode.

• #11208 Fixed the issue of abnormal data statistics for LwM2M clients.

• #11211 HTTP API DELETE operations on non-existent resources now consistently returns 404.

• #11214 Fixed a bug where node configuration may fail to synchronize correctly when the node joins the cluster.

• #11229 Fixed an issue that prevented plugins from starting/stopping after changing configuration via emqx ctl conf load.

• #11237 The headers default value in /prometheus API should be a map instead of a list.

• #11250 Fixed a bug when the order of MQTT packets withing a WebSocket packet will be reversed.

• #11271 Ensured that the range of all percentage type configurations is from 0% to 100% in the REST API and configuration. For example, sysom.os.sysmem_high_watermark=101% is invalid now.

• #11272 Fixed a typo in the log, where an abnormal PUBREL packet was mistakenly referred to as pubrec.

• #11281 Restored support for the special $queue/ shared subscription topic prefix.

• #11294 Fixed emqx ctl cluster join, leave, and status commands.

• #11306 Fixed rule action metrics inconsistency where dropped requests were not accounted for.

• #11309 Improved startup order of EMQX applications. Simplified build scripts and improved code reuse.

• #11322 Added support for importing additional configurations from EMQX backup file (emqx ctl import command):

  • rule_engine (previously not imported due to the bug)
  • topic_metrics (previously not implemented)
  • slow_subs (previously not implemented).

• #10645 Changed health check for Oracle Database, PostgreSQL, MySQL and Kafka Producer data bridges to ensure target table/topic exists.

• #11107 MongoDB bridge health check now returns the failure reason.

• #11139 The Redis bridge has been refactored into its own Erlang application to improve the code structure and to make it easier to maintain.

• #11151 The MySQL bridge has been refactored into its own Erlang application to improve the code structure and to make it easier to maintain.

• #11163 Hid topology.pool_size in MondoDB bridges and fixed it to 1 to avoid confusion.

• #11175 Now when using a nonexistent hostname for connecting to MySQL, a 400 error is returned rather than 503 in the REST API.

• #11198 Fixed global rebalance status evaluation on replicant nodes. Previously, /api/v5/load_rebalance/global_status API method could return incomplete results if handled by a replicant node.

• #11223 In InfluxDB bridging, mixing decimals and integers in a field may lead to serialization failure in the Influx Line Protocol, resulting in the inability to write to the InfluxDB bridge (when the decimal point is 0, InfluxDB mistakenly interprets it as an integer).

  See also: InfluxDB v2.7 Line-Protocol.

• #11225 The username field in PostgreSQL/Timescale/MatrixDB bridges configuration is now a required one.

• #11242 Restarted emqx_ee_schema_registry when a node joins a cluster. As emqx_ee_schema_registry uses Mria tables, a node joining a cluster needs to restart this application in order to start relevant Mria shard processes, ensuring a correct behaviour in Core/Replicant mode.

• #11266 Fixed and improved support for TDengine insert syntax:

  1. Added support for inserting into multi-table in the template.

     For example:

     insert into table_1 values (${ts}, '${id}', '${topic}') table_2 values (${ts}, '${id}', '${topic}')

  2. Added support for mixing prefixes/suffixes and placeholders in the template.

     For example:

     insert into table_${topic} values (${ts}, '${id}', '${topic}')

     Note: This is a breaking change. Previously, the values of string type were quoted automatically, but now they must be quoted explicitly.

     For example:

     insert into table values (${ts}, '${a_string}')

• #11307 Fixed check for table existence to return a more friendly message in the Oracle bridge.

• #11316 Fixed Pool Size value not being considered in Oracle Bridge.

• #11326 Fixed return error checking on table validation in the Oracle bridge.

5.1.0

Release Date: 2023-06-21

Enhancements

• #11035 Upgraded Cassandra driver to avoid username and password leakage in data bridge logs.
• #10584 Added log level configuration to SSL communication
• #10678 Optimized counter increment calls to avoid work if increment is zero.
• #10690 Added a retry mechanism to webhook bridge that attempts to improve throughput. This optimization retries request failures without blocking the buffering layer, which can improve throughput in situations of high messaging rate.
• #10702 Introduced a more straightforward configuration option keepalive_multiplier and deprecate the old keepalive_backoff configuration. After this enhancement, EMQX checks the client's keepalive timeout status period by multiplying the "Client Requested Keepalive Interval" with keepalive_multiplier.
• #10698 Optimized memory usage when accessing the configuration during runtime.
• #10778 Refactored Pulsar Producer bridge to avoid leaking resources in case bridge crashed during initialization phase.
• #10813 Refactored Kafka Producer and Consumer bridges to avoid leaking resources in case bridge crashed during initialization phase.
• #10858 A new utility function timezone_to_offset_seconds/1 has been added to the rule engine SQL language. This function converts a timezone string (for example, "+02:00", "Z" and "local") to the corresponding offset in seconds.
• #10841 Added a schema validation to ensure message key is not empty when "key_dispatch" strategy is selected in Kafka and Pulsar Producer bridges.
• #10754 The MQTT bridge has been enhanced to utilize connection pooling and leverage available parallelism, substantially improving throughput. As a consequence, single MQTT bridge now uses a pool of clientids to connect to the remote broker.
• #10782 Added a new deliver_rate option to the retainer configuration, which can limit the maximum delivery rate per session in the retainer.
• #10877 Upgraded RocketMQ driver to enhance security for sensitive data.
• #10598 Provided a callback method of Unary type in ExProto to avoid possible message disorder issues.
• #10895 Refactored most of the bridges to avoid resource leaks in case bridge crashed during initialization phase.
• #10790 Optimized access to configuration in runtime by reducing overhead of reading configuration per zone.
• #10892 Added the requirement for setting SID or Service Name in Oracle Database bridge creation.
• #10910 The data bridge resource option auto_restart_interval was deprecated in favor of health_check_interval, and request_timeout was renamed to request_ttl. Also, the default request_ttl value went from 15 seconds to 45 seconds. The previous existence of both auto_restart_interval and health_check_interval was a source of confusion, as both parameters influenced the recovery of data bridges under failures. An inconsistent configuration of those two parameters could lead to messages being expired without a chance to retry. Now, health_check_interval is used both to control the interval of health checks that may transition the data bridge into disconnected or connecting states, as well as recovering from disconnected.
• #10929 Upgraded Erlang/OTP to 25.3.2-1.
• #10909 Removed the deprecated HTTP APIs for gateways.
• #10908 Refactored the RocketMQ bridge to avoid resources leaks in case bridge crashed during initialization phase.
• #10924 Refactored Influxdb bridge connector to avoid resource leaks in case bridge crashed during initialization phase.
• #10944 Improved the GCP PubSub bridge to avoid a potential issue that the bridge could fail to send messsages after node restart.
• #10933 Added support for configuring TCP keep-alive in MQTT/TCP and MQTT/SSL listeners.
• #10948 Added live_connections field for some HTTP APIs, i.e:
  • /monitor_current, /monitor_current/nodes/{node}
  • /monitor/nodes/{node}, /monitor
  • /node/{node}, /nodes
• #10941 Improved the collection speed of Prometheus metrics when setting prometheus.vm_dist_collector=disabled and metric erlang_vm_statistics_run_queues_length_total is renamed to erlang_vm_statistics_run_queues_length
• #10985 Renamed emqx ctl command cluster_call to conf cluster_sync. The old command cluster_call is still a valid command, but not included in usage info.
• #10988 Improved log security when data bridge creation fails to ensure sensitive data is always obfuscated.
• #10926 Allowed enable as well as enabled as the state flag for listeners. Prior to this change, listener can be enable/disabled by setting the true or false on the enabled config. This is slightly different naming comparing to other state flags in the system. Now the enable flag is added as an alias in listener config.
• #10970 A query_mode parameter has been added to the Kafka producer bridge. This parameter allows you to specify if the bridge should use the asynchronous or synchronous mode when sending data to Kafka. The default is asynchronous mode.
• #10676 Added CLI commands emqx ctl export and emqx ctl import for importing/exporting configuration and user data. This allows exporting configurations and built-in database data from a running EMQX cluster and importing them into the same or another running EMQX cluster.
• #11003 Added an option to configure TCP keepalive in Kafka bridge.
• #10961 Added support for unlimited max connections for gateway listeners by allowing infinity as a valid value for the max_connections field in the configuration and HTTP API.
• #11019 Improved log security for JWT, now it will be obfuscated before print.
• #11024 Added a small improvement to reduce the chance of seeing the connecting state when creating/updating a Pulsar Producer bridge.
• #11034 Hid the broker config and changed the broker.shared_subscription_strategy to mqtt.shared_subscription_strategy as it belongs to mqtt.
• #11045 The listener's authentication and zone related apis have been officially removed in version 5.1.0.
• #11062 Renamed config log.file.to to log.file.path.

Bug Fixes

• #11018 Fixed multiple issues with the Stomp gateway, including:

  • Fixed an issue where is_superuser was not working correctly.
  • Fixed an issue where the mountpoint was not being removed in message delivery.
  • After a message or subscription request fails, the Stomp client should be disconnected immediately after replying with an ERROR message.

• #11051 Added validation to ensure that certificate depth (listener SSL option) is a non negative integer.

• #10563 Corrected an issue where the no_local flag was not functioning correctly in subscription.

• #10653 Stored gateway authentication TLS certificates and keys in the data directory to fix the problem of memory leakage.

• #10682 Fixed the timestamp for the will message is incorrectly assigned at the session creation time, now this timestamp is the disconnected time of the session.

• #10701 RPM package for Amazon Linux 2 did not support TLS v1.3 as it was assembled with Erlang/OTP built with openssl 1.0.

• #10677 Fixed an issue in the Rule API where attempting to delete a non-existent rule resulted in a 404 HTTP error code response.

• #10715 Support for getting the client certificate in the client.connected hook. Previously, this data was removed after the connection was established to reduce memory usage.

• #10737 Fixed the issue where the HTTP API interface of Gateway cannot handle ClientIDs with special characters, such as: !@#$%^&*()_+{}:"<>?/.

• #10809 Addressed ** ERROR ** Mnesia post_commit hook failed: error:badarg error messages happening during node shutdown or restart. Mria pull request: https://github.com/emqx/mria/pull/142

• #10807 The debug-level logs related to license checks will no longer be printed. These logs were generated too frequently and could interfere with log recording.

• #10818 Fixed emqx_ctl traces command error where the traces start command in the emqx_mgmt_cli module was not working properly with some filters.

• #10600 Deleted emqx_statsd application.

• #10820 Fixed the issue where newly added nodes in the cluster would not apply the new license after a cluster license update and would continue to use the old license. Sometimes the new node must start with a outdated license. e.g. use emqx-operator deployed and needed to scale up after license expired. At the time the cluster's license key already updated by API/CLI, but the new node won't use it.

• #10851 Obfuscated sensitive data in the bad API logging.

• #10884 Fixed an issue where trying to get rule info or metrics could result in a crash when a node is joining a cluster.

• #10887 Fixed a potential issue where requests to bridges might take a long time to be retried. This only affected low throughput scenarios, where the buffering layer could take a long time to detect connectivity and driver problems.

• #10878 Fixed a vulnerability in the RabbitMQ bridge, which could potentially expose passwords to log files.

• #10871 Fixed an issue where the Dashboard shows that the connection still exists after a CoAP connection is disconnected, but deletion and message posting requests do not take effect.

• #10880 Added a new REST API POST /clients/kickout/bulk for kicking out multiple clients in bulk.

• #10913 Fixed an issue where the plugin status REST API of a node would still include the cluster node status after the node left the cluster.

• #10923 Fixed a race-condition in channel info registration. Prior to this fix, when system is under heavy load, it might happen that a client is disconnected (or has its session expired) but still can be found in the clients page in dashboard. One of the possible reasons is a race condition fixed in this PR: the connection is killed in the middle of channel data registration.

• #10930 Added a schema validation for duration data type to avoid invalid values. Before this fix, it was possible to use absurd values in the schema that would exceed the system limit, causing a crash.

• #10952 Disallow enabling fail_if_no_peer_cert in listener SSL options if verify = verify_none is set. Setting fail_if_no_peer_cert = true and verify = verify_none caused connection errors due to incompatible options. This fix validates the options when creating or updating a listener to avoid these errors.

  Note: any old listener configuration with fail_if_no_peer_cert = true and verify = verify_none that was previously allowed will fail to load after applying this fix and must be manually fixed.

• #10951 Fixed the issue in MQTT-SN gateway when the mountpoint did not take effect on message publishing.

• #10943 Deprecated UDP mcast mechanism for cluster discovery. This feature has been planed for deprecation since 5.0 mainly due to the lack of actual production use. This feature code is not yet removed in 5.1, but the document interface is demoted.

• #10902 Avoid syncing cluser.hocon file from the nodes running a newer version than the self-node. During cluster rolling upgrade, if an older version node has to restart due to whatever reason, if it copies the cluster.hocon file from a newer version node, it may fail to start. After this fix, the older version node will not copy the cluster.hocon file from a newer, so it will use its own cluster.hocon file to start.

• #10967 Fixed error message formatting in rebalance API: previously they could be displayed as unclear dumps of internal Erlang structures. Added wait_health_check option to node evacuation CLI and API. This is a time interval when the node reports "unhealthy status" without beginning actual evacuation. We need this to allow a Load Balancer (if any) to remove the evacuated node from balancing and not forward (re)connecting clients to the evacuated node.

• #10911 The error message and log entry that appear when one tries to create a bridge with a name the exceeds 255 bytes is now easier to understand.

• #10983 Fixed the issue when mqtt clients could not connect over TLS if the listener was configured to use TLS v1.3 only. The problem was that TLS connection was trying to use options incompatible with TLS v1.3.

• #10977 Fixed the delay in updating subscription count metric and corrected configuration issues in Stomp gateway.

• #10950 Fixed the issue where the enable_qos option does not take effect in the MQTT-SN gateway.

• #10999 Changed schema validation for Kafka fields 'Partition Count Refresh Interval' and 'Offset Commit Interval' to avoid accepting values larger then maximum allowed.

• #10997 The ClickHouse bridge had a problem that could cause messages to be dropped when the ClickHouse server is closed while sending messages even when the request_ttl is set to infinity. This has been fixed by treating errors due to a closed connection as recoverable errors.

• #10994 Redacted proxy-authorization headers as used by HTTP connector to avoid leaking secrets into log files.

• #10996 For any unknown HTTP/API request, the default response is a 404 error rather than the dashboard's index.html.

• #11005 Fixed the issue where the method field cannot be correctly printed in the trace logs of AuthN HTTP.

• #11006 Fixed QUIC listeners's default cert file paths. Prior to this change, the default cert file paths are prefixed with environment variable ${EMQX_ETC_DIR} which were not interpolated before used in QUIC listeners.

• #10998 Do not allow batch_size option for MongoDB bridge resource. MongoDB connector currently does not support batching, the batch_size config value is forced to be 1 if provided.

• #10955 Fixed the issue in MQTT-SN gateway where deleting Predefined Topics configuration does not work.

• #11025 Fixed a case_clause error that could arise in race conditions in Pulsar Producer bridge.

• #11030 Improved error messages when a validation error occurs while using the Listeners HTTP API.

• #11033 Deprecated the mountpoint field in AuthenticateRequest in ExProto gateway. This field was introduced in e4.x, but in fact, in e5.0 we have provided gateway.exproto.mountpoint for configuration, so there is no need to override it through the Authenticate request.

  Additionally, updates the default value of subscriptions_max, inflight_max, mqueue_max to infinity.

• #11040 Fixed a health check issue for Kafka Producer that could lead to loss of messages when the connection to Kafka's brokers were down.

• #11038 Fixed a health check issue for Pulsar Producer that could lead to loss of messages when the connection to Pulsar's brokers were down.

• #11042 Fixed crash on REST API GET /listeners when listener's max_connections is set to a string.

• #11028 Disallowed using multiple TLS versions in the listener config that include tlsv1.3 but exclude tlsv1.2. Using TLS configuration with such version gap caused connection errors. Additionally, drop and log TLS options that are incompatible with the selected TLS version(s).

  Note: any old listener configuration with the version gap described above will fail to load after applying this fix and must be manually fixed.

• #11031 Fixed credential validation when creating bridge and checking status for InfluxDB Bridges.

• #11056 Fixed the issue where newly created listeners sometimes do not start properly. When you delete a system default listener and add a new one named 'default', it will not start correctly.

  • Fixed the bug where configuration failure on certain nodes can cause Dashboard unavailability.

• #11070 Fixed the problem that the cluster.autoclean configuration item does not take effect.

• #11092 and #11100 Fixed problem when replicat nodes were unable to connect to the core node due to timeout in mria_lb:core_nodes() call. Relevant mria pull request: https://github.com/emqx/mria/pull/143

5.0.26

Release Date: 2023-05-29

Enhancements

• #10584 Add log level configuration to SSL communication

• #10702 Introduce a more straightforward configuration option keepalive_multiplier and deprecate the old keepalive_backoff configuration. After this enhancement, EMQX checks the client's keepalive timeout status period by multiplying the "Client Requested Keepalive Interval" with keepalive_multiplier.

• #10713 We hide the request_timeout in resource_option of the webhook to keep it consistent with the http request_timeout of the webhook. From now on, when configuring a webhook through API or configuration files, it is no longer necessary to configure the request_timeout of the resource. Only configuring the http request_timeout is sufficient, and the request_timeout in the resource will automatically be consistent with the http request_timeout.

• #10511 Improve the security and privacy of some resource logs by masking sensitive information in the data.

• #10678 Optimized counter increment calls to avoid work if increment is zero.

• #10690 Added a retry mechanism to webhook bridge that attempts to improve throughput.

  This optimization retries request failures without blocking the buffering layer, which can improve throughput in situations of high messaging rate.

• #10698 Optimize memory usage when accessing the configuration during runtime.

Bug Fixes

• #10340 Fixed the issue that could lead to crash logs being printed when stopping EMQX via systemd.

• #10563 Corrected an issue where the no_local flag was not functioning correctly.

• #10600 Deleted emqx_statsd application.

• #10653 Store gateway authentication TLS certificates and keys in the data directory.

• #10677 In Rule API, reapond with 404 HTTP error code when trying to delete a rule that does not exist.

• #10682 Fix the timestamp for the will message is incorrectly assigned at the session creation time, now this timestamp is the disconnected time of the session.

• #10701 RPM package for Amazon Linux 2 did not support TLS v1.3 as it was assembled with Erlang/OTP built with openssl 1.0.

• #10715 Postpone trimming the connection information structure until after client.connected hooks have been executed. These hooks once again have access to the client's peer certificate.

• #10717 Fixed an issue where the buffering layer processes could use a lot of CPU when inflight window is full.

• #10724 A summary has been added for all endpoints in the HTTP API documentation (accessible at "http://emqx_host_name:18083/api-docs").

• #10726 Validate Health Check Interval and Auto Restart Interval against the range from 1ms to 1 hour.

• #10728 Fixed an issue where the rule engine was unable to access variables exported by FOREACH in the DO clause.

  Given a payload: {"date": "2023-05-06", "array": ["a"]}, as well as the following SQL statement:

  FOREACH payload.date as date, payload.array as elem
  DO date, elem
  FROM "t/#"

  Prior to the fix, the date variable exported by FOREACH could not be accessed in the DO clause of the above SQL, resulting in the following output for the SQL statement: [{"elem": "a","date": "undefined"}]. After the fix, the output of the SQL statement is: [{"elem": "a","date": "2023-05-06"}]

• #10737 Fix the issue where the HTTP API interface of Gateway cannot handle ClientIDs with special characters, such as: !@#$%^&*()_+{}:"<>?/.

• #10742 Check the correctness of the rules before saving the authorization file source. Previously, Saving wrong rules could lead to restart failure.

• #10743 Fixes an issue where trying to get a bridge info or metrics could result in a crash when a node is joining a cluster.

• #10746 Add missing support of the event $events/delivery_dropped into the rule engine test API rule_test.

• #10747 Refactor date and time functions, format_date and date_to_unix_ts, in the rule engine to fix the implementation problem.

• #10755 Fixed data bridge resource update race condition.

  In the 'delete + create' process for EMQX resource updates, long bridge creation times could cause dashboard request timeouts. If a bridge resource update was initiated before completion of its creation, it led to an erroneous deletion from the runtime, despite being present in the config file.

  This fix addresses the race condition in bridge resource updates, ensuring the accurate identification and addition of new resources, maintaining consistency between runtime and configuration file statuses.

• #10760 Fix Internal Error 500 that occurred sometimes when bridge statistics page was updated while a node was (re)joining the cluster.

• #10761 Fixing the issue where the default value of SSL certificate for Dashboard Listener was not correctly interpolated, which caused HTTPS to be inaccessible when verify_peer and cacertfile were using the default configuration.

• #10785 Ensure EMQX_LOG_DIR is set by Windows boot script.

  The environment variable EMQX_LOG_DIR was missing in v5.0.25, caused EMQX Windows package fail to boot unless set by sysadmin.

• #10801 Avoid duplicated percent decode the topic name in API /topics/{topic} and /topics.

• #10809 Address ** ERROR ** Mnesia post_commit hook failed: error:badarg error messages happening during node shutdown or restart. Mria pull request: https://github.com/emqx/mria/pull/142

• #10817 Fix the error of not being able to configure auto_restart_interval as infinity

• #10818 Fixing emqx_ctl traces command.

• #10820 In case the cluster updated license before the new node join in. The new node will not apply the updated license. After this change, the new joined node will use the cluster's license key.

  Sometimes the new node must start with a outdated license. e.g. use emqx-operator deployed and needed to scale up after license expired. At the time the cluster's license key already updated by API/CLI, but the new node won't use it.

• #10833 Only include enabled authenticators and authorizers in telemetry report, not all of them.

• #10851 Obfuscated sensitive data in the bad API logging.

5.0.25

Release Date: 2023-05-12

Enhancements

• #10568 Add shutdown counter information to emqx ctl listeners command

• #10571 Do not emit useless crash report when EMQX stops. Previously, when EMQX (and emqx_topic_metrics in particular) stopped and removed underlying tables, some messages were still being handled and crashed.

• #10588 Increase the time precision of trace logs from second to microsecond. For example, change from 2023-05-02T08:43:50+00:00 to 2023-05-02T08:43:50.237945+00:00.

• #10623 Renamed max_message_queue_len to max_mailbox_size in the force_shutdown configuration. Old name is kept as an alias, so this change is backward compatible.

• #10417 Improve get config performance by eliminating temporary references.

• #10525 Reduce resource usage per MQTT packet handling.

• #10528 Reduce memory footprint in hot code path.

• #10573 Improved performance of Webhook bridge when using synchronous query mode. This also should improve the performance of other bridges when they are configured with no batching.

• #10591 Improve the configuration of the limiter.

  • Simplify the memory representation of the limiter configuration.
  • Make sure the node-level limiter can really work when the listener's limiter configuration is omitted.

• #10625 Simplify limiter configuration.

  • Reduce the complexity of the limiter's configuration. e.g. now users can use limiter.messages_rate = 1000/s to quickly set the node-level limit for the message publish.
  • Update the configs/limiter API to suit this refactor.

Bug Fixes

• #10548 Fixed a race condition in the HTTP driver that would result in an error rather than a retry of the request. Related fix in the driver: https://github.com/emqx/ehttpc/pull/45

• #10556 Wrap potentially sensitive data in emqx_connector_http if Authorization headers are being passed at initialization.

• #10659 Fix the issue where emqx cannot start when sysmon.os.mem_check_interval is disabled.

5.0.24

Release Date: 2023-04-26

Enhancements

• #10457 Deprecates the integration with StatsD.

  There seemd to be no user using StatsD integration, so we have decided to hide this feature for now. We will either remove or revive it based on requirements in the future.

• #10458 Set the level of plugin configuration options to low level, in most cases, users only need to manage plugins on the dashboard without the need for manual modification, so we lowered the level.

• #10491 Rename etcd.ssl to etcd.ssl_options to keep all of SSL options consistent in the configuration file.

• #10512 Improved the storage format of Unicode characters in data files, Now we can store Unicode characters normally. For example: "SELECT * FROM "t/1" WHERE clientid = "-测试专用-""

• #10487 Optimize the instance of limiter for whose rate is infinity to reduce memory and CPU usage.

• #10490 Remove the default limit of connect rate which used to be 1000/s

Bug Fixes

• #10407 Improve 'emqx_alarm' performance by using Mnesia dirty operations and avoiding unnecessary calls from 'emqx_resource_manager' to reactivate alarms that have been already activated. Use new safe 'emqx_alarm' API to activate/deactivate alarms to ensure that emqx_resource_manager doesn't crash because of alarm timeouts. The crashes were possible when the following conditions co-occurred:

  • a relatively high number of failing resources, e.g. bridges tried to activate alarms on re-occurring errors;
  • the system experienced a very high load.

• #10420 Fix HTTP path handling when composing the URL for the HTTP requests in authentication and authorization modules.

  • Avoid unnecessary URL normalization since we cannot assume that external servers treat original and normalized URLs equally. This led to bugs like #10411.
  • Fix the issue that path segments could be HTTP encoded twice.

• #10422 Fixed a bug where external plugins could not be configured via environment variables in a lone-node cluster.

• #10448 Fix a compatibility issue of limiter configuration introduced by v5.0.23 which broke the upgrade from previous versions if the capacity is infinity.

  In v5.0.23 we have replaced capacity with burst. After this fix, a capacity = infinity config will be automatically converted to equivalent burst = 0.

• #10449 Validate the ssl_options and header configurations when creating authentication http (authn_http). Prior to this, incorrect ssl configuration could result in successful creation but the entire authn being unusable.

• #10455 Fixed an issue that could cause (otherwise harmless) noise in the logs.

  During some particularly slow synchronous calls to bridges, some late replies could be sent to connections processes that were no longer expecting a reply, and then emit an error log like:

  2023-04-19T18:24:35.350233+00:00 [error] msg: unexpected_info, mfa: emqx_channel:handle_info/2, line: 1278, peername: 172.22.0.1:36384, clientid: caribdis_bench_sub_1137967633_4788, info: {#Ref<0.408802983.1941504010.189402>,{ok,200,[{<<"cache-control">>,<<"max-age=0, ...">>}}

  Those logs are harmless, but they could flood and worry the users without need.

• #10462 Deprecate config broker.shared_dispatch_ack_enabled. This was designed to avoid dispatching messages to a shared-subscription session which has the client disconnected. However since v5.0.9, this feature is no longer useful because the shared-subscrption messages in a expired session will be redispatched to other sessions in the group. See also: https://github.com/emqx/emqx/pull/9104

• #10463 Improve bridges API error handling. If Webhook bridge URL is not valid, bridges API will return '400' error instead of '500'.

• #10484 Fix the issue that the priority of the configuration cannot be set during rolling upgrade. For example, when authorization is modified in v5.0.21 and then upgraded v5.0.23 through rolling upgrade, the authorization will be restored to the default.

• #10495 Add the limiter API /configs/limiter which was deleted by mistake back.

• #10500 Add several fixes, enhancements and features in Mria:

  • protect mria:join/1,2 with a global lock to prevent conflicts between two nodes trying to join each other simultaneously Mria PR
  • implement new function mria:sync_transaction/4,3,2, which blocks the caller until a transaction is imported to the local node (if the local node is a replicant, otherwise, it behaves exactly the same as mria:transaction/3,2) Mria PR
  • optimize mria:running_nodes/0Mria PR
  • optimize mria:ro_transaction/2 when called on a replicant node Mria PR.

• #10518 Add the following fixes and features in Mria:

  • call mria_rlog:role/1 safely in mria_membership to ensure that mria_membership gen_server won't crash if RPC to another node fails Mria PR
  • Add extra field to ?rlog_sync table to facilitate extending this functionality in future Mria PR.

5.0.23

Release Date: 2023-04-18

Enhancements

• #10156 Change the priority of the configuration:

  1. If it is a new installation of EMQX, the priority of configuration is ENV > emqx.conf > HTTP API.
  2. If EMQX is upgraded from an old version (i.e., the cluster-override.conf file still exists in EMQX's data directory), then the configuration priority remains the same as before. That is, HTTP API > ENV > emqx.conf.

  Deprecated data/configs/local-override.conf.

  Stabilizing the HTTP API for hot updates.

• #10354 More specific error messages when configure with bad max_heap_size value. Log current value and the max value when the message_queue_too_long error is thrown.

• #10359 Metrics now are not implicitly collected in places where API handlers don't make any use of them. Instead, a separate backplane RPC gathers cluster-wide metrics.

• #10373 Deprecate the trace.payload_encode configuration. Add payload_encode=[text,hidden,hex] option when creating a trace via HTTP API.

• #10389 Unify the config formats for cluster.core_nodes and cluster.statics.seeds. Now they both support formats in array ["emqx1@127.0.0.1", "emqx2@127.0.0.1"] or semicolon-separated string "emqx1@127.0.0.1,emqx2@127.0.0.1".

• #10391 Hide a large number of advanced options to simplify the configuration file.

  That includes rewrite, topic_metric, persistent_session_store, overload_protection, flapping_detect, conn_congestion, stats,auto_subscribe, broker_perf, shared_subscription_group, slow_subs, ssl_options.user_lookup_fun and some advance items in node and dashboard section, #10358, #10381, #10385.

• #10392 A new function to convert a formatted date to an integer timestamp has been added: date_to_unix_ts/3

• #10404 Change the default queue mode for buffer workers to memory_only. Before this change, the default queue mode was volatile_offload. When under high message rate pressure and when the resource is not keeping up with such rate, the buffer performance degraded a lot due to the constant disk operations.

• #10426 Optimize the configuration priority mechanism to fix the issue where the configuration changes made to etc/emqx.conf do not take effect after restarting EMQX.

  More introduction about the new mechanism: Configure Override Rules

• #10376 Simplify the configuration of the limiter feature and optimize some codes

  • Rename message_in to messages
  • Rename bytes_in to bytes
  • Use burst instead of capacity
  • Hide non-importance fields
  • Optimize limiter instances in different rate settings

• #10430 Simplify the configuration of the retainer feature.

  • Mark flow_control as non-importance field.

Bug Fixes

• #10369 Fix error in /api/v5/monitor_current API endpoint that happens when some EMQX nodes are down.

  Prior to this fix, sometimes the request returned HTTP code 500 and the following message:

  {"code":"INTERNAL_ERROR","message":"error, badarg, [{erlang,'++',[{error,nodedown},[{node,'emqx@10.42.0.150'}]], ...

• #10410 Fix config check failed when gateways are configured in emqx.conf. This issue was first introduced in v5.0.22 via #10278, the boot-time config check was missing.

5.0.22

Release Date: 2023-04-13

Enhancements

• #10077 Add support for QUIC TLS password protected certificate file.

• #10128 Add support for OCSP stapling for SSL MQTT listeners.

• #10164 Add CRL check support for TLS MQTT listeners.

• #10206 Decouple the query mode from the underlying call mode for buffer workers.

  Prior to this change, setting the query mode of a resource such as a bridge to sync would force the buffer to call the underlying connector in a synchronous way, even if it supports async calls.

• #10207 Use 'label' from i18n file as 'summary' in OpenAPI spec.

• #10210 Unregister Mnesia post commit hook when Mria is being stopped. This fixes hook failures occasionally occurring on stopping/restarting Mria.

  Mria PR

• #10224 Add the option to customize clusterIP in Helm chart, so that a user may set it to a fixed IP.

• #10263 Add command 'eval-ex' for Elixir expression evaluation.

• #10278 Refactor the directory structure of all gateways.

• #10306 Add support for async query mode for most bridges.

  Before this change, some bridges (Cassandra, MongoDB, MySQL, Postgres, Redis, RocketMQ, TDengine) were only allowed to be created with a sync query mode.

• #10318 Now, the rule engine language's FROM clause supports both strings enclosed in double quotes (") and single quotes (').

• #10336 Add /rule_engine API endpoint to manage configuration of rule engine.

Bug Fixes

• #10145 Fix bridges API to report error conditions for a failing bridge as status_reason. Also when creating an alarm for a failing resource we include this error condition with the alarm's message.

• #10154 Change the default resume_interval for bridges and connectors to be the minimum of health_check_interval and request_timeout / 3. Also exposes it as a hidden configuration to allow fine tuning.

  Before this change, the default values for resume_interval meant that, if a buffer ever got blocked due to resource errors or high message volumes, then, by the time the buffer would try to resume its normal operations, almost all requests would have timed out.

• #10172 Fix the incorrect default ACL rule, which was:

  {allow, {username, "^dashboard?"}, subscribe, ["$SYS/#"]}.

  However, it should use {re, "^dashboard$"} to perform a regular expression match:

  {allow, {username, {re,"^dashboard$"}}, subscribe, ["$SYS/#"]}.

• #10174 Upgrade library esockd from 5.9.4 to 5.9.6. Fix an unnecessary error level logging when a connection is closed before proxy protocol header is sent by the proxy.

• #10195 Add labels to API schemas where description contains HTML and breaks formatting of generated documentation otherwise.

• #10196 Use lower-case for schema summaries and descritptions to be used in menu of generated online documentation.

• #10209 Fix bug where a last will testament (LWT) message could be published when kicking out a banned client.

• #10211 Hide broker.broker_perf config and API documents. The two configs route_lock_type and trie_compaction are rarely used and requires a full cluster restart to take effect. They are not suitable for being exposed to users. Detailed changes can be found here: https://gist.github.com/zmstone/01ad5754b9beaeaf3f5b86d14d49a0b7/revisions

• #10225 Allow installing a plugin if its name matches the beginning of another (already installed) plugin name. For example: if plugin "emqx_plugin_template_a" is installed, it must not block installing plugin "emqx_plugin_template".

• #10226 Don't crash on validation error in /bridges API, return 400 instead.

• #10237 Ensure we return 404 status code for unknown node names in /nodes/:node[/metrics|/stats] API.

• #10242 Fixed a log data field name clash. Piror to this fix, some debug logs may report a wrong Erlang PID which may affect troubleshooting session takeover issues.

• #10251 Consider bridges referenced in FROM rule clauses as dependencies.

  Before this fix, when one tried to delete an ingress rule referenced in an action like select * from "$bridges/mqtt:ingress", the UI would not trigger a warning about dependent rule actions.

• #10257 Fixed the issue where auto_observe was not working in LwM2M Gateway.

  Before the fix, OBSERVE requests were sent without a token, causing failures that LwM2M clients could not handle.

  After the fix, LwM2M Gateway can correctly observe the resource list carried by client, furthermore, unknown resources will be ignored and printing the following warning log:

  2023-03-28T18:50:27.771123+08:00 [warning] msg: ignore_observer_resource, mfa: emqx_lwm2m_session:observe_object_list/3, line: 522, peername: 127.0.0.1:56830, clientid: testlwm2mclient, object_id: 31024, reason: no_xml_definition

• #10286 Enhance logging behaviour during boot failure. When EMQX fails to start due to corrupted configuration files, excessive logging is eliminated and no crash dump file is generated.

• #10297 Keeps eval command backward compatible with v4 by evaluating only Erlang expressions, even on Elixir node. For Elixir expressions, use eval-ex command.

• #10300 Fixed an issue where a build made with Elixir could not receive uploaded plugins until the plugins folder was created manually to receive the uploaded files.

• #10313 Ensure that when the core or replicant node starting, the cluster-override.conf file is only copied from the core node. Previously, when sorting nodes by startup time, the core node may have copied this file from the replicant node.

• #10314 Fix /monitor_current API so that it only looks at the current node. Fix /stats API to not crash when one or more nodes in the cluster are down.

• #10315 Fix crash checking limit and page parameters in /mqtt/delayed/messages API call.

• #10317 Do not expose listener level authentications before extensive verification.

• #10323 For security reasons, the value of the password field in the API examples is replaced with ******.

• #10327 Don't increment 'actions.failed.unknown' rule metrics counter upon receiving unrecoverable bridge errors. This counter is displayed on the dashboard's rule overview tab ('Action statistics' - 'Unknown'). The fix is only applicable for synchronous bridges, as all rule actions for asynchronous bridges are counted as successful (they increment 'actions.success' which is displayed as 'Action statistics' - 'Success').

5.0.21

Release Date: 2023-03-24

Enhancements

• #10022 Start releasing Rocky Linux 9 (compatible with Enterprise Linux 9) and MacOS 12 packages

• #10139 Add extraVolumeMounts to EMQX Helm Chart, it will have the ability to mount the user-own files into the EMQX instance, for example, ACL rule files as mentioned in #9052

  Done of #10116

• #9893 When connecting with the flag clean_start=false, EMQX will filter out messages that published by banned clients. Previously, the messages sent by banned clients may still be delivered to subscribers in this scenario.

• #9986 For helm charts, add MQTT ingress bridge; and removed stale mgmt references.

• #10123 Improve the performance of /bridges API. Earlier, when the number of nodes in the cluster was large or the node was busy, the API may have a request timeout.

• #9998 Redact the HTTP request body in the authentication error logs for security reasons.

Bug Fixes

• #10013 Fix return type structure for error case in API schema for /gateways/:name/clients. u

• #10014 In dashboard API for /monitor(_current)/nodes/:node return 404 instead of 400 if node does not exist.

• #10026 Metrics are now only exposed via the /bridges/:id/metrics endpoint. Metrics are no longer returned in other API operations such as getting the list of all bridges, or in the response when a bridge has been created.

• #10027 Allow setting node name from EMQX_NODE__NAME when running in docker. Prior to this fix, only EMQX_NODE_NAME is allowed.

• #10050 Ensure Bridge API returns 404 status code consistently for resources that don't exist.

• #10052 Improve daemon mode startup failure logs.

  Before this change, it was difficult for users to understand the reason for EMQX 'start' command failed to boot the node. The only information they received was that the node did not start within the expected time frame, and they were instructed to boot the node with 'console' command in the hope of obtaining some logs. However, the node might actually be running, which could cause 'console' mode to fail for a different reason.

  With this new change, when daemon mode fails to boot, a diagnosis is issued. Here are the possible scenarios:

  • If the node cannot be found from ps -ef, the user is instructed to find information in log files erlang.log.*.
  • If the node is found to be running but not responding to pings, the user is advised to check if the host name is resolvable and reachable.
  • If the node is responding to pings, but the EMQX app is not running, it is likely a bug. In this case, the user is advised to report a Github issue.

• #10055 Fix: configuration parameter mqtt.max_awaiting_rel had no effect.

• #10056 Fix /bridges API status code.

  • Return 400 instead of 403 in case of removing a data bridge that is dependent on an active rule.
  • Return 400 instead of 403 in case of calling operations (start|stop|restart) when Data-Bridging is not enabled.

• #10066 Improve error messages for /briges_probe and [/node/:node]/bridges/:id/:operation API calls to make them more readable. And set HTTP status code to 400 instead of 500.

• #10074 Check if type in PUT /authorization/sources/:type matches type given in body of request.

• #10079 Fix description of shared_subscription_strategy.

• #10085 Consistently return 404 for all requests on non existent source in /authorization/sources/:source[/*].

• #10098 A crash with an error in the log file that happened when the MongoDB authorization module queried the database has been fixed.

• #10100 Fix channel crash for slow clients with enhanced authentication. Previously, when the client was using enhanced authentication, but the Auth message was sent slowly or the Auth message was lost, the client process would crash.

• #10107 For operations on bridges API if bridge-id is unknown we now return 404 instead of 400. Also a bug was fixed that caused a crash if that was a node operation. Additionally we now also check if the given bridge is enabled when doing the cluster operation start . Affected endpoints:

  • [cluster] /bridges/:id/:operation,
  • [node] /nodes/:node/bridges/:id/:operation, where operation is one of [start|stop|restart]. Moreover, for a node operation, EMQX checks if node name is in our cluster and return 404 instead of 501.

• #10117 Fix an error occurring when a joining node doesn't have plugins that are installed on other nodes in the cluster. After this fix, the joining node will copy all the necessary plugins from other nodes.

• #10118 Fix problems related to manual joining of EMQX replicant nodes to the cluster. Previously, after manually executing joining and then leaving the cluster, the replicant node can only run normally after restarting the node after joining the cluster again.

  Mria PR

• #10119 Fix crash when statsd.server is set to an empty string.

• #10124 The default heartbeat period for MongoDB has been increased to reduce the risk of too excessive logging to the MongoDB log file.

• #10130 Fix garbled config display in dashboard when the value is originally from environment variables. For example, env EMQX_STATSD__SERVER='127.0.0.1:8124' . /bin/emqx start results in unreadable string (not '127.0.0.1:8124') displayed in Dashboard's Statsd settings page. Related PR: HOCON#234.

• #10132 Fix some error logs generated by systemctl stop emqx command. Prior to the fix, the command was not stopping jq and os_mon applications properly.

• #10144 Add -setcookie emulator flag when invoking emqx ctl to prevent problems with emqx cli when home directory is read only. Fixes #10142.

• #10157 Fixed default rate limit configuration not being applied correctly when creating a new listener.

5.0.20

Release Date: 2023-03-10

Enhancements

• #10059 Errors returned by rule engine API are formatted in a more human readable way rather than dumping the raw error including the stacktrace.

Bug Fixes

• #10032 When resources on some nodes in the cluster are still in the 'initializing/connecting' state, the bridges/ API will crash due to missing Metrics information for those resources. This fix will ignore resources that do not have Metrics information.

• #10044 Fix node information formatter for stopped nodes in the cluster. The bug was introduced by v5.0.18.

• #10054 Fix the problem that the obfuscated password is used when using the /bridges_probe API to test the connection in Data-Bridge.

• #10058 Deprecate unused QUIC TLS options. Only following TLS options are kept for the QUIC listeners:

  • cacertfile
  • certfile
  • keyfile
  • verify

• #10076 Fix webhook bridge error handling: connection timeout should be a retriable error. Prior to this fix, connection timeout was classified as unrecoverable error and led to request being dropped.

• #10078 Fix an issue that invalid QUIC listener setting could casue segfault.

• #10084 Fix problem when joining core nodes running different EMQX versions into a cluster.

  Mria PR

• #10086 Upgrade HTTP client ehttpc to 0.4.7. Prior to this upgrade, HTTP clients for authentication, authorization and webhook may crash if body is empty but content-type HTTP header is set. For more details see ehttpc PR#44.

5.0.19

Release Date: 2023-03-01

Bug Fixes

• #10032 When the resource manager is busy trying to establish a connection with the remote, the resource might yet lack any metrics information. Prior to this fix, the bridges/ API handler crashed in such circumstances.

• #10037 Fix Swagger API doc rendering crash. In version 5.0.18, a bug was introduced that resulted in duplicated field names in the configuration schema. This, in turn, caused the Swagger schema generated to become invalid.

• #10041 For influxdb bridge, added integer value placeholder annotation hint to write_syntax documentation. Also supported setting a constant value for the timestamp field.

• #10042 Improve behavior of the replicant nodes when the core cluster becomes partitioned (for example when a core node leaves the cluster). Previously, the replicant nodes were unable to rebalance connections to the core nodes, until the core cluster became whole again. This was indicated by the error messages: [error] line: 182, mfa: mria_lb:list_core_nodes/1, msg: mria_lb_core_discovery divergent cluster.

  Mria PR

• #10043 Fixed two bugs introduced in v5.0.18.

  • The environment varialbe SSL_DIST_OPTFILE was not set correctly for non-boot commands.
  • When cookie is overridden from environment variable, EMQX node is unable to start.

• #10044 Fix node information formatter for stopped nodes in the cluster.

5.0.18

Release Date: 2023-02-24

Enhancements

• #10019 Add low level tuning settings for QUIC listeners.

• #9213 Add pod disruption budget to helm chart

• #9949 QUIC transport Multistreams support and QUIC TLS cacert support.

• #9967 New common TLS option 'hibernate_after' to reduce memory footprint per idle connecion, default: 5s.

Bug Fixes

• #10009 Validate bytes param to GET /trace/:name/log to not exceed signed 32bit integer.

• #10015 To prevent errors caused by an incorrect EMQX node cookie provided from an environment variable, we have implemented a fail-fast mechanism. Previously, when an incorrect cookie was provided, the command would still attempt to ping the node, leading to the error message 'Node xxx not responding to pings'. With the new implementation, if a mismatched cookie is detected, a message will be logged to indicate that the cookie is incorrect, and the command will terminate with an error code of 1 without trying to ping the node.

• #10020 Fix bridge metrics when running in async mode with batching enabled (batch_size > 1).

• #10021 Fix error message when the target node of emqx_ctl cluster join command is not running.

• #9939 Allow 'emqx ctl cluster' command to be issued before Mnesia starts. Prior to this change, EMQX replicant could not use manual discovery strategy. Now it's possible to join cluster using 'manual' strategy.

• #9958 Fix bad http response format when client ID is not found in clients APIs

• #9961 Avoid parsing config files for node name and cookie when executing non-boot commands in bin/emqx.

• #9974 Report memory usage to statsd and prometheus using the same data source as dashboard. Prior to this fix, the memory usage data source was collected from an outdated source which did not work well in containers.

• #9978 Fixed configuration issue when choosing to use SSL for a Postgres connection (authn, authz and bridge). The connection could fail to complete with a previously working configuration after an upgrade from 5.0.13 to newer EMQX versions.

• #9997 Fix Swagger API schema generation. deprecated metadata field is now always boolean, as Swagger specification suggests.

5.0.17

Release Date: 2023-02-13

Enhancements

• #9802 Support HAProxy protocol for HTTP API.

• #9871 Allow the placeholder to be anywhere in the topic for authz rules. e.g: {allow, {username, "who"}, publish, ["t/foo${username}boo/${clientid}xxx"]}.

• #9910 Add start operation to bridges API to allow manual reconnect after failure.

• #9917 Stop building -alpine docker image because it's size is larger than the regular one based on debian slim

• #9930 Expose the stats live_connections.count and live_connections.max to Prometheus.

• #9936 Disable disksup (part of os_mon) in releases by default, no warnings are issued when a disk error occurs.

• #9954 Improve bridge performance

Bug fixes

• #9864 Fix the exclusive topics aren't removed when the session has already been cleaned.

• #9875 Return 400 if a broken plugin package is uploaded from HTTP API, also cleanup if plugin is not accepted.

• #9916 Fix MQTT bridge fails to verify TLS wildcard server certificate.

• #9922 Fix the issue with the bridge resource buffer where it might become stuck if enough async queries fill the inflight window full before failing with retryable errors.

• #9923 Fix REPORT_CB/2 CRASH error logs when errors happen during boot-up or shutdown.

• #9938 Report some egress MQTT bridge errors as recoverable, and thus retryable.

• #9946 Add back reconnect_interval as deprecated field for MQTT bridge. The field was removed from v5.0.16/e5.0.0 by mistake, caused new version unable to start on old config. Now it's added back as deprecated (config value is ignored if provided).

• #9951 Propagate errors from operations (start|stop|restart) on bridges API if called for all nodes.

• #9952 Disallow subscribing with QoS 2 for ingress MQTT bridges. Allow user to configure clean_start option for ingress MQTT bridges, however.

5.0.16

Release Date: 2023-02-02

Bug fixes

• #9824 The topics/{topic} API endpoint would return 500 - Internal Error if a topic had multiple routes. This is fixed by returning a list of routes.

• #9832 Improve error log when bridge in 'sync' mode timed out to get response.

• #9834 Allow mqtt.idle_timeout to be set to infinity

• #9839 Make sure that the content of an authorization header that users have specified for a webhook bridge is not printed to log files.

• #9884 Do not resume all buffer workers on successful health check of any individual resource. Previously after any successful healthcheck, all buffer workers (for all resources) were resumed

5.0.15

Release Date: 2023-01-20

Enhancements

• #9569 Refactor /authorization/sources/built_in_database/ by adding rules/ to the path.

• #9585 /bridges_probe API endpoint to test params for creating a new data bridge.

• #9586 Basic auth is no longer allowed for API calls, must use API key instead.

• #9628 Expose additional resource configuration parameters: start_after_created and start_timeout.

• #9722 Add the following configuration options for Pushing metrics to Prometheus Push Gateway:

  • headers: Allows custom HTTP request headers.
  • job_name: allows to customize the name of the Job pushed to Push Gateway.

• #9725 Remove the config auto_reconnect from the emqx_authz, emqx_authn and data-bridge componets. This is because we have another config with similar functions: resource_opts.auto_restart_interval。

  The functions of these two config are difficult to distinguish, which will lead to confusion. After this change, auto_reconnect will not be configurable (always be true), and the underlying drivers that support this config will automatically reconnect the abnormally disconnected connection every 2s.

  And the config resource_opts.auto_restart_interval is still available for user. It is the time interval that emqx restarts the resource when the connection cannot be established for some reason.

• #9736 Refactor of /bridges API to make it more consistent with other APIs:

  • bridge enable/disable is now done via the endpoint /bridges/{id}/enable/[true,false]
  • /bridges/{id}/operation/{operation} endpoints are now /bridges/{id}/{operation}
  • metrics are moved out from the GET /bridges/{id} response and can now be fetched via /bridges/{id}/metrics
  • the bridges/{id}/reset_metrics endpoint is now /bridges/{id}/metrics/reset

• #9774 Add a password complexity requirement when adding or modifying Dashboard users via the API. Now password must contain at least 2 of alphabetic, numeric and special characters, and must be 8 to 64 characters long.

Bug fixes

• #9626 Return authorization settings with default values. The authorization cache is enabled by default, but due to the missing default value in GET response of /authorization/settings, it seemed to be disabled from the dashboard.

• #9680 Fix the problem that username and password authentication is mandatory in Influxdb v1 write API.

• #9726 Client fuzzy search API results were missing information which could tell if more results are available in the next pages, this is now fixed by providing hasnext flag in the response.

• #9735 Password information has been removed from information log messages for http, ldap, mongo, mqtt, mysql, pgsql and redis.

• #9748 Listeners not configured with max_connections will cause the cluster /listeners API to return 500 error.

• #9749 In some cases search APIs could respond with an incorrect count value in the metadata, that is usually much bigger than expected, this is now fixed.

• #9750 Reload overriding configs after boot. Prior to this change, two configs were allow to change from dashboard, but will not take effect after reboot:

  • Logging (such as level)
  • Prometheus configs

• #9751 Fix that obsoleted cert file will not be deleted after the listener is updated/deleted

• #9763 Fix an authentication exception when password is not provided

• #9765 Parse decimals as password from environment variable overrides correctly. Prior to this change, config values for passwords are not allowed to be decimals. e.g. EMQX_FOOBAR__PASSWORD=12344 or emqx.foobar.password=1234 would result in a type check error, unless quoted as: EMQX_FOOBAR__PASSWORD='"12344"' or emqx.foobar.password="1234". After this fix, the value does not have to be quoted.

• #9769 Fix Erlang shell prompt version prefix. e5.0.15 -> v5.0.15

• #9780 When creating disk queue directory for resource worker, substitute ':' with '-' in worker id.

• #9781 Trace files were left on a node when creating a zip file for download. They are now removed when the file is sent. Also, concurrent downloads will no longer interfere with each other.

• #9785 Stop authentication hook chain if emqx_authentication provides a definitive result.

• #9787 Fix a compatible problem for the webhook bridge configuration which was created before the v5.0.12.

5.0.14

Release Date: 2023-01-11

Enhancements

• https://github.com/emqx/emqx/pull/8329 The MongoDB library has been upgraded to support MongoDB 5.1+

• https://github.com/emqx/emqx/pull/9593 Obfuscated sensitive data in the response when querying bridges information by API.

• https://github.com/emqx/emqx/pull/9614 Make it possible to configure host:port from environment variables without quotes. Prior to this change, when overriding a host:port config value from the environment variable, one has to quote it as: env EMQX_BRIDGESMQTTXYZSERVER='"localhost:1883"'. Now it's possible to set it without quote as env EMQX_BRIDGESMQTTXYZSERVER='localhost:1883'.

• https://github.com/emqx/emqx/pull/9642 Deprecates enable_batch and enable_queue options for bridges/resources. After this change, queuing is always enabled for bridges, and batching is controlled by the batch_size option: batch_size > 1 means batching will be enabled.

• https://github.com/emqx/emqx/pull/9671 Implement sliding window average metrics.

• https://github.com/emqx/emqx/pull/9674 Made rule engine behavior more consistent with bridge behavior regarding metrics: if a rule engine is disabled, its metrics are now reset.

• https://github.com/emqx/emqx/pull/9675 HTTP client library ehttpc upgraded from 0.4.2 to 0.4.3. Library eredis_cluster which manages clients to Redis clusters upgraded from 0.7.1 to 0.7.5.

• https://github.com/emqx/emqx/pull/9713 Introduce api_key.bootstrap_file to initialize the api key at boot time. Deprecate dashboard.bootstrap_users_file. Limit the maximum number of api keys to 100 instead of 30.

Bug fixes

• https://github.com/emqx/emqx/pull/8648 When deleting a non-existing bridge the server gave a successful response. This has been fixed so that the server instead gives an error response when the user attempts to delete a non-existing bridge.

• https://github.com/emqx/emqx/pull/9637 Fix the expiry_interval fields of the clients HTTP API to measure in seconds.

• https://github.com/emqx/emqx/pull/9638 Fix the problem of data loss and bad match when the MySQL driver is disconnected.

• https://github.com/emqx/emqx/pull/9641 Fix an issue where testing the GCP PubSub could leak memory and an issue where its JWT token would fail to refresh a second time.

• https://github.com/emqx/emqx/pull/9642 Fix some issues that could lead to wrong bridge metrics. Fix an issue that could lead to message loss and wrong metrics with the Kafka Producer bridge when Kafka or the connection to it is down. Fix some issues that could lead to the same message being delivered more than once when using batching for bridges and when the batch was retried.

• https://github.com/emqx/emqx/pull/9667 Remove the possibility to set clientid for /publish and /publish/bulk HTTP APIs. This is to reduce the risk of security confusion.

• https://github.com/emqx/emqx/pull/9687 Fix the problem that sending messages to data-bridges failed because of incorrect handling of some data-bridges without local_topic field configured. Before this change, if some bridges have configured the local_topic field but others have not, a function_clause error will occur when forwarding messages to the data-bridges.

• https://github.com/emqx/emqx/pull/9689 Fix handling of HTTP authorization result when a request failure (e.g.: HTTP resource is down) would cause a function_clause error.

• https://github.com/emqx/emqx/pull/9703 Set the default value of the qos field of the HTTP API /clients/:clientid/subscribe to 0. Before this fix, the qos field has no default value, which leads to a function_clause error when querying this API.

• https://github.com/emqx/emqx/pull/9705 Remove the default value of Webhook. Before this repair, the default value of the body field of Webhook is ${payload}, but there is no payload field in the available fields of other events except the message publishing in the rule, so in this case, the webhook will send a string with the message body as "undefined" to the HTTP service. This fix removes the default value of the body field. When the body field is not configured, Webhook will send all available fields of the current event in the format of a JSON object.

• https://github.com/emqx/emqx/pull/9712 Fixed the problem of '404 Not Found' when calling the HTTP API '/clients/:clientid/subscribe/bulk' from the plug-ins and data-bridges on handling the 'client.connected' event.

• https://github.com/emqx/emqx/pull/9714 Fix /mqtt/auto_subscribe API's bad swagger schema, and make sure swagger always checks if the schema is correct.

• https://github.com/emqx/emqx/pull/9716 MQTT bridge config compatibility fix. The config created before v5.0.12 may encounter a compatibility issue after upgrading to v5.0.13.

• https://github.com/emqx/emqx/pull/9717 Prior to this fix, if it always times out when trying to connect a bridge server, it's not possible to change other configs even when the bridge is disabled.

5.0.13

Release Date: 2022-12-27

Enhancements

• Add limiter update API #9133.

• Avoid creating temporary zip files when syncing data directory during cluster startup #9429.

• Refactor: move /mqtt/sys_topics to generic /configs/sys_topics #9511.

• Refactor: use POST not PUT for /users/{name}/change_pwd #9533.

• Add compression functions zip, gzip, zip_compress in Rule-Engine and corresponding decompression functions #9573.

• Return 204 instead of 200 for PUT /authenticator/:id #9434.

• Added the option to customize the clientid prefix of egress MQTT bridges. #9609

• Ensure the default expiration time of banned is large enough #9599.

Bug fixes

• Trigger message.dropped hook when QoS2 message is resend by client with a same packet id, or 'awaiting_rel' queue is full #9487.

• Fix shared subscription 'sticky' strategy #9578. Prior to this change, a 'sticky' subscriber may continue to receive messages after unsubscribing.

• Add check to ensure that a given key is among the prepared statements on query in the mysql connector #9571.

• Fix password leak to logs for connectors #9608.

5.0.12

Release Date: 2022-12-14

Highlights

• This version included a refactoring of MQTT bridge config. The older version config file created from v5.0.11 or earlier will be converted to according to the new schema. Please note though, the request body of /bridges API to configure MQTT brdige is changed in a incompatible way.
• Start releasing packages for Apple M1/M2 (MacOS-12).
• Start releasing packages for Amazon Linux 2 (e.g. emqx-5.0.12-amzn2-amd64.rpm).
• Retained message index performance improvement.

Enhancements

• Disable global garbage collection by node.global_gc_interval = disabled #9418。

• Improve the CLI to avoid waste atom table when typing erros #9416.

• Start building MacOS packages for Apple Silicon hadrdware #9423.

• Remove support for setting shared subscriptions using the non-standard $queue feature #9412. Shared subscriptions are now part of the MQTT spec. Use $share instead.

• Refactor authn API by replacing POST /authentication/{id}/move with PUT /authentication/{id}/position/{position}. #9419. Same is done for /listeners/{listener_id}/authentication/id/....

• Redesign /rules API to make metrics a dedicated resources rather than being included with every response #9461.

• Add more PSK ciphers support #9505.

• Improve emqx_retainer write performance: get rid of transactions on write #9372.

• HTTP client library ehttpc upgraded from 0.4.0 to 0.4.2 #9520.

• Add handshake_timeout option to MQTT SSL listener #9502.

• Upgrade dashboard to v1.1.3.

• Users can define the externalTrafficPolicy of service in EMQX Helm Chart #9527.

• Return 204 instead of 200 for POST /gateway/lwm2m/clients/{clientid}/{read,write,observe} #9480.

• Make possible to create an authentication entirely from environment variable #9547. As an example, one can now enable MySQL auth with: env EMQX_AUTHENTICATION__1='{mechanism="password_based",backend="mysql",server="localhost:3306",database="emqx",username="emqx",password="******",query="SELECT password_hash,salt FROM mqtt_user WHERE username=${username} LIMIT 1",enable=true}'. Prior to this change, overrides only work on top of existing authentication, for example, if there is already MySQL auth configured in emqx.conf but we want to disable it, we can do it with env EMQX_AUTHENTICATION__1__ENABLE=false.

• Start building packages for Amazon Linux 2 #9537.

Bug fixes

• Fix that the obsolete SSL files aren't deleted after the ExHook config update #9432.

• Fix doc and schema for /trace API #9468.

• Return 404 for /telemetry/data in case it's disabled #9464.

• Fix some potential MQTT packet parse errors #9477.

• Fixed EMQX Helm Chart deployment error #9509.

  • Fixed the Discovery error: no such service error occurred during helm chart deployment, resulting in an abnormal discovery of cluster nodes.
  • Fixed issue that caused EMQX Helm Chart to fail when modifying some of EMQX's configuration items via environment variables.

• Fix shadowing 'client.authenticate' callbacks by emqx_authenticator. Now emqx_authenticator passes execution to the further callbacks if none of the authenticators matches #9496.

• Return 400 if query param node is not a known node in /trace/:id/download?node={node} #9478.

• POST /traces to return 409 in case of duplicate #9494.

• Fix bridging function, when both ingress and egress bridges are configured, egress bridge does not work #9523.

• Fix EMQX Helm Chart using incorrect secret values when custom credentials are provided #9536.

5.0.11

Release Date: 2022-11-27

Enhancements

• Security enhancement for retained messages #9326. The retained messages will not be published if the publisher client is banned.

• Security enhancement for the subscribe API #9355.

• Enhance the banned feature #9367. Now the corresponding session will be kicked when client is banned by clientid.

• Redesign /gateways API 9364. Use PUT /gateways/{name} instead of POST /gateways, gateway gets 'loaded' automatically if needed. Use PUT /gateways/{name}/enable/{true|false} to enable or disable gateway. No more DELETE /gateways/{name}.

• Support statsd {tags: {"user-defined-tag" = "tag-value"} configure and improve stability of emqx_statsd #9363.

• Improve node name generation rules to avoid potential atom table overflow risk #9387.

• Set the default value for the maximum level of a topic to 128 #9406.

• Keep MQTT v5 User-Property pairs from bridge ingested MQTT messsages to bridge target #9398.

Bug fixes

• Fix ssl.existingName option of helm chart not working #9307.

• Fix create trace sometime failed by end_at time has already passed. #9303

• Return 404 for status of unknown authenticator in /authenticator/{id}/status #9328.

• Fix that JWT ACL rules are only applied if an exp claim is set #9368.

• Fix that /configs/global_zone API cannot get the default value of the configuration #9392.

• Fix mountpoint not working for will-msg #9399.

5.0.10

Release Date: 2022-11-09

Release had to be recreated due to an issue in GitHub action which failed to upload/publish packages. Previous commit: 34a6c6c88

Enhancements

• Improve /nodes API responsiveness #9221.

• Improve the integration of the banned and the delayed feature #9326. Now when publishing a delayed message will check first if its source client is banned, if true, this publish will be ignored.

• Update gen_rpc library to version 3.0 #9187.

• Improve memory usage on core nodes when bootstrapping a replicant #9236.

• Improve stability of Prometheus Push Gateway and log errors when POST fails #9235.

• Now it is possible to opt out VM internal metrics in prometheus stats #9222. When system load is high, reporting too much metrics data may cause the prometheus stats API timeout.

• Improve security when converting types such as binary lists to atom types #9279, #9286.

• Add /trace/:name/log_detail HTTP API to return trace file's size and mtime #9152.

• Add /status HTTP API endpoint to api documentation #9230.

• Binary packages for all platforms are now built on Erlang/OTP version 24.3.4.2 #9293.

Bug fixes

• Fix error log message when mechanism is missing in authentication config #8924.

• Fix HTTP 500 issue when unknown status parameter is used in /gateway API call #9225.

• Fixed the HTTP response status code for the /status endpoint #9211. Before the fix, it always returned 200 even if the EMQX application was not running. Now it returns 503 in that case.

• Fix message delivery related event encoding #9228. This bug was introduced in v5.0.9. For Rule-Engine's input events like $events/message_delivered and $events/message_dropped, if the message was delivered to a shared-subscription, the encoding (to JSON) of the event will fail.

• Fix bad HTTP response status code for /gateways API, when Gateway name is unknown, it should return 404 instead of 400 #9268.

• Fix incorrect topic authorize checking of delayed messages #9290. Now will determine the actual topic of the delayed messages, e.g. $delayed/1/t/foo will be treated as t/foo in authorize checks.

• Add property code to error response for /authentication/sources/:type 9299.

• Align documentation for /authentication/sources with what we actually send 9299.

• Fix query string parameter 'node' to /configs resource being ignored, return 404 if node does not exist #9310.

• Avoid re-dispatching shared-subscription session messages when a session is kicked or taken-over (to a new session) #9123.

5.0.9

Release Date: 2022-10-24

Enhancements

• Add cert_common_name and cert_subject placeholder support for authz_http and authz_mongo #8973.

• Use milliseconds internally in emqx_delayed to store the publish time, improving precision #9060.

• More rigorous checking of flapping to improve stability of the system #9136.

• No message(s) echo for the message publish APIs #9155. Prior to this fix, the message publish APIs (api/v5/publish and api/v5/publish/bulk) echos the message back to the client in HTTP body. This change fixed it to only send back the message ID.

Bug fixes

• Check ACLs for last will testament topic before publishing the message #8930.

• Fix GET /listeners API crash when some nodes (in a cluster) is still loading the configs #9002.

• Fix empty variable interpolation in authentication and authorization #8963. Placeholders for undefined variables are rendered now as empty strings and do not cause errors anymore.

• Fix the latency statistics error of the slow subscription stats #8986. Prior to this change when stats_type is internal or response, the begin time stamp was taken at wrong precision.

• Fix shared subscription message re-dispatches #9104.

  • When discarding QoS 2 inflight messages, there were excessive logs
  • For wildcard deliveries, the re-dispatch used the wrong topic (the publishing topic, but not the subscribing topic), caused messages to be lost when dispatching.

• Upgrade http client gun from 1.3.7 to 1.3.9 Prior to this fix, long-lived HTTPS connections for HTTP auth or webhook integrations may stall indefinitely, causing massive timeouts for HTTP requests.

5.0.8

Release Date: 2022-09-17

Changes Worth Mentioning

Enhancements

• An independent RPC implementation is used between nodes to forward shared subscription messages instead of Erlang's own RPC to reduce the cluster pressure when the shared subscription load is high. #8893
• Print a warning message when boot with the default (insecure) Erlang cookie. #8905
• The configuration in local-override.conf will not allow synchronous updates in the cluster to the entire cluster at runtime. #8851
• Add POST /listeners interface for creating listeners. #8876
• Change /gateway API path to plural form. #8823
• The exp, nbf and iat claims in JWT authentication support non-integer timestamps. #8867
• Improve the request performance of ExProto and gRPC Server. #8866

Bug Fixes

• Fix the issue that password authentication using Redis as the data source directly terminates the authentication when no authentication data is retrieved. #8934
• Fix inaccurate delayed publish due to OS time changes. #8926
• Fix the issue that EMQX could not be started after disabling the retained message feature. #8911
• Fix the slow response of updating the configuration when a node in the cluster is down. #8857
• Fix the issue that the authorization would terminate the execution of the client.authorize hook when no rules were matched. #8780
• Fix the issue that Payload must be configured in MQTT Bridge. #8949
• Fix the issue that the log directory cannot be configured through environment variables. #8892
• Fix the issue that the CoAP gateway introduced an extra / prefix when parsing topics. #8658
• Fix the issue that MQTT Bridge returned the content of the TLS file in the API response. #8872, #8958
• Fix the issue that client authentication failure would trigger the release of will messages. #8887
• Fix ExProto's imperfect Keep Alive check mechanism that could cause the client to never expire. #8866

Dependency Upgrades

• grpc-erl upgrades from 0.6.6 to 0.6.7, #8866

5.0.7

Release Date: 2022-09-01

Enhancements

• Simplify TLS cipher suite configuration
• Add confirmation before listener closes on Dashboard
• Unify the configuration of TLS on Dashboard
• Supports viewing EMQX version and node role on Dashboard overview page
• Increase restrictions on plugin file types

Bug fixes

• Fix the issue that Mria transactions of replicated nodes could not be executed when nodes of mixed versions formed a cluster
• Fix the issue that the Authorization settings page of Dashboard could not display data
• Fix the issue that the monitoring topic data could not be reset on Dashboard

Other changes

• Remove will message related fields in the response of the client query interface

5.0.6

Release Date: 2022-08-22

Bug fixes

• Fix incorrect display of node status on Dashboard

5.0.4

Release Date: 2023-05-26

Enhancements

• #10389 Unified the configuration formats for cluster.core_nodes and cluster.statics.seeds. Now they both support formats in array ["emqx1@127.0.0.1", "emqx2@127.0.0.1"] and the comma-separated string "emqx1@127.0.0.1,emqx2@127.0.0.1".

• #10392 Introduced a new function to convert a formatted date to an integer timestamp: date_to_unix_ts/3.

  date_to_unix_ts(TimeUnit, FormatString, InputDateTimeString)

• #10426 Optimized the configuration priority mechanism to fix the issue where the configuration changes made to etc/emqx.conf do not take effect after restarting EMQX.

  More information about the new mechanism: Configure Override Rules

• #10457 Deprecated the integration with StatsD.

• #10458 Set the level of plugin configuration options to low, users usually manage the plugins through the dashboard, rarely modify them manually, so we lowered the level.

• #10491 Renamed etcd.ssl to etcd.ssl_options to keep all SSL options consistent in the configuration file.

• #10512 Improved the storage format of Unicode characters in data files, Now we can store Unicode characters. For example: SELECT * FROM "t/1" WHERE clientid = "-测试专用-".

• #10568 Added shutdown_count printout to emqx ctl listeners command.

• #10588 Increased the time precision of trace logs from second to microsecond. For example, change from 2023-05-02T08:43:50+00:00 to 2023-05-02T08:43:50.237945+00:00.

• #10623 Renamed max_message_queue_len to max_mailbox_size in the force_shutdown configuration. The old name is kept as an alias, so this change is backward compatible.

• #10713 Hide the resource_option.request_timeout of the webhook and it will use the value of http request_timeout.

• #10075 Added node rebalance/node evacuation functionality. See also: EIP doc

• #10378 Implemented Pulsar Producer Bridge and only producer role is supported now.

• #10408 Introduced 3 built-in functions in the rule engine SQL-like language for creating values of the MongoDB date type.

• #10409 #10337 Supported Protocol Buffers and Apache Avro schemas in Schema Registry.

• #10425 Implemented OpenTSDB data bridge.

• #10498 Implemented Oracle Database Bridge.

• #10560 Added enterprise data bridge for Apache IoTDB.

• #10417 Improved get config items performance by eliminating temporary references.

• #10430 Simplified the configuration of the retainer feature. Marked flow_control as a non-importance field.

• #10511 Improved the security and privacy of some resource logs by masking sensitive information in the log.

• #10525 Reduced resource usage per MQTT packet handling.

• #10528 Reduced memory footprint in hot code path. The hot path includes the code that is frequently executed in core functionalities such as message handling, connection management, authentication, and authorization.

• #10591 #10625 Improved the configuration of the limiter.

  • Reduced the complexity of the limiter's configuration.

  • Updated the configs/limiter API to suit this refactor.

  • Reduced the memory usage of the limiter configuration.

• #10487 Optimized the instance of limiter for whose rate is infinity to reduce memory and CPU usage.

• #10490 Removed the default limit of connect rate which used to be 1000/s.

• #10077 Added support for QUIC TLS password-protected certificate file.

Bug Fixes

• #10340 Fixed the issue that could lead to crash logs being printed when stopping EMQX via systemd.

• #10369 Fixed error in /api/v5/monitor_current API endpoint that happens when some EMQX nodes are down.

  Prior to this fix, sometimes the request returned HTTP code 500 and the following message:

  {"code":"INTERNAL_ERROR","message":"error, badarg, [{erlang,'++',[{error,nodedown},[{node,'emqx@10.42.0.150'}]], ...

• #10407 Fixed the crash issue of the alarm system.

  • Leverage Mnesia dirty operations and circumvent extraneous calls to enhance 'emqx_alarm' performance.

  • Use 'emqx_resource_manager' for reactivating alarms that have already been triggered.

  • Implement the newly developed, fail-safe 'emqx_alarm' API to control the activation and deactivation of alarms, thus preventing 'emqx_resource_manager' from crashing due to alarm timeouts.

  • The alarm system is susceptible to crashing under these concurrent conditions:

    • A significant number of resources fail, such as when bridges continuously attempt to trigger alarms due to recurring errors.

    • The system is under an extremely high load.

• #10420 Fixed HTTP path handling when composing the URL for the HTTP requests in authentication and authorization modules.

  • Avoid unnecessary URL normalization since we cannot assume that external servers treat original and normalized URLs equally. This led to bugs like #10411.

  • Fixed the issue that path segments could be HTTP encoded twice.

• #10422 Fixed a bug where external plugins could not be configured via environment variables in a lone-node cluster.

• #10448 Fixed a compatibility issue of limiter configuration introduced by e5.0.3 which broke the upgrade from previous versions if the capacity is infinity.

  In e5.0.3 we have replaced capacity with burst. After this fix, a capacity = infinity config will be automatically converted to equivalent burst = 0.

• #10462 Deprecated config broker.shared_dispatch_ack_enabled. This was designed to avoid dispatching messages to a shared-subscription session that has the client disconnected. However, since e5.0.0, this feature is no longer helpful because the shared-subscription messages in an expired session will be redispatched to other sessions in the group. See also: https://github.com/emqx/emqx/pull/9104 .

• #10463 Improved bridges API error handling. If Webhook bridge URL is not valid, the bridges API will return '400' error instead of '500'.

• #10484 Fixed the issue that the priority of the configuration cannot be set during the rolling upgrade. For example, when authorization is modified in e5.0.2 and then upgraded e5.0.3 through the rolling upgrade, the authorization will be restored to the default.

• #10495 Added the limiter API /configs/limiter which was deleted by mistake back.

• #10500 Added several fixes, enhancements, and features in Mria:

  • Protect mria:join/1,2 with a global lock to prevent conflicts between two nodes trying to join each other simultaneously Mria PR

  • Implement new function mria:sync_transaction/4,3,2, which blocks the caller until a transaction is imported to the local node (if the local node is a replicant, otherwise, it behaves exactly the same as mria:transaction/3,2) Mria PR

  • Optimize mria:running_nodes/0 Mria PR

  • Optimize mria:ro_transaction/2 when called on a replicant node Mria PR.

• #10518 Added the following fixes and features in Mria:

  • Call mria_rlog:role/1 safely in mria_membership to ensure that mria_membership gen_server won't crash if RPC to another node fails Mria PR

  • Add an extra field to ?rlog_sync table to facilitate extending this functionality in future Mria PR.

• #10556 Wrapped potentially sensitive data in emqx_connector_http if Authorization headers are being passed at initialization.

• #10571 Stopped emitting useless crash report when EMQX stops.

• #10659 Fixed the issue where EMQX cannot start when sysmon.os.mem_check_interval is disabled.

• #10717 Fixed an issue where the buffering layer processes could use a lot of CPU when inflight window is full.

• #10724 A summary has been added for all endpoints in the HTTP API documentation (accessible at "http://<emqx_host_name>:18083/api-docs").

• #10726 Health Check Interval and Auto Restart Interval now support the range from 1ms to 1 hour.

• #10728 Fixed an issue where the rule engine was unable to access variables exported by FOREACH - DO clause.

  Given a payload: {"date": "2023-05-06", "array": ["a"]}, as well as the following SQL statement:

  FOREACH payload.date as date, payload.array as elem DO date, elem FROM "t/#" -- {"date": "2023-05-06", "array": ["a"]}

  Prior to the fix, the date variable exported by FOREACH could not be accessed in the DO clause of the above SQL, resulting in the following output for the SQL statement: [{"elem": "a","date": "undefined"}].

• #10742 Correctness check of the rules is enforced before saving the authorization file source. Previously, Saving wrong rules could lead to EMQX restart failure.

• #10743 Fixed an issue where trying to get bridge info or metrics could result in a crash when a node is joining a cluster.

• #10755 Fixed data bridge resource update race condition.

  In the 'delete + create' process for EMQX resource updates, long bridge creation times could cause dashboard request timeouts. If a bridge resource update was initiated before completion of its creation, it led to an erroneous deletion from the runtime, despite being present in the config file.

  This fix addresses the race condition in bridge resource updates, ensuring the accurate identification and addition of new resources, and maintaining consistency between runtime and configuration file statuses.

• #10761 Fixed the issue where the default value of SSL certificate for Dashboard Listener was not correctly interpolated, which caused HTTPS to be inaccessible when verify_peer and cacertfile were using the default configuration.

• #10672 Fixed the issue where the lack of a default value for ssl_options in listeners results in startup failure. For example, such command(EMQX_LISTENERS__WSS__DEFAULT__BIND='0.0.0.0:8089' ./bin/emqx console) would have caused a crash before.

• #10738 TDEngine data bridge now supports "Supertable" and "Create Tables Automatically". Before this fix, an insert with a supertable in the template will fail, like this:

  • insert into ${clientid} using msg TAGS (${clientid}) values (${ts},${msg}).

• #10746 Add missing support of the event $events/delivery_dropped into the rule engine test API rule_test.

• #10747 Ported some time formating fixes in Rule-Engine functions from version 4.4.

• #10760 Fix "internal error 500" when getting bridge statistics page while a node is joining the cluster.

• #10801 Avoid double percent-decode for topic name in API /topics/{topic} and /topics.

• #10817 Fix a config value handling for bridge resource option auto_restart_interval, now it can be set to infinity.

5.0.4

Release Date: 2022-07-28

Enhancements

• Rules in data Integration support paging and searching. Note that the GET /rules API will return the page meta information after the update, i.e. {"data": [RuleObj1, RuleObj2], "meta": {"count": 2, "limit": 100, "page": 1}}. #8472
• Improve the health check of WebHook in data integration, if TLS is enabled, it will now check if the TLS handshake was successful. #8443
• Falls back to using Mnesia to persist sessions when RocksDB is unavailable. #8528
• Support for updating thresholds of alarms at runtime via the HTTP API. #8532
• Log trace will show the detailed authentication process. #8554
• Supports listening on IPv6 addresses, for example: [::1]:1883 or ::1:1883. #8547
• Updated the Listener API's request and response formats to align with the behavior of other APIs. This will introduce some incompatible updates, see #8571。
• Dashboard will prompt to change the default password.
• Optimize Dashboard's overview page for rules and data bridges of data integration.
• Add result statistics for data integration rules on Dashboard.
• Optimize the charts in the Dashboard homepage.
• Add MQTT 5.0 subscription options display to subscription list on Dashboard.

Bug fixes

• Fix the issue that when the log type format is set to json, the configuration of the maximum length of a single log is invalid. #8518
• Fix the issue that jq in data integration cannot be used when the path of the EMQX installation directory contains spaces. #8455
• Fix the issue that super user does not take effect. #8452
• Aligned system topic format for stats with metrics. #8464
• Fix the issue that the creation time of the rules in the data integration was not persistent, causing it to be updated to the EMQX startup time every time. #8443
• Fix an issue where the cluster-override.conf file would be emptied when there was an error updating the configuration via the HTTP API, causing all modified configurations to be lost. #8443
• Fix the issue that the Sentinel field was not required to be set when using Redis sentinel mode in authentication and authorization. #8458
• Fix formatting errors in OpenAPI documentation. #8517
• Fix the issue that multilingual hook extensions might not be dispatched in the order in which client events were fired. #8530
• Fix authentication placeholders cert_subject and cert_common_name not being available. #8531
• Fix TCP connection process leak in WebHook. ehttpc#34, #8580
• Fix CLI not printing listeners that only listen on ports. #8547
• Fix incorrect TLS field checking in JWKS authentication. #8458
• Fix listener API not returning connection information on all nodes in the cluster. #8538
• Fix the issue that replicant nodes might not receive Mnesia events, causing operations such as configuration updates to fail. #8502

5.0.3

Release Date: 2023-05-08

Enhancements

• #10128 Add support for OCSP stapling for SSL MQTT listeners.

• #10156 Change the configuration overlay order:

  If it is a new installation of EMQX, emqx.conf + Environment variables overlays on top of API Updated Configs (cluster.hocon)

  If EMQX is upgraded from an older version (i.e., the cluster-override.conf file still exists in EMQX's data directory), then it’s the same as before, that is cluster-override.conf overlays on top of emqx.conf + Environment variables.

  Please note that data/configs/cluster-override.conf is considered deprecated. After upgrade, you are encouraged to update emqx.conf to delete configs which are overridden by cluster-override.conf and move the configs in cluster-override.conf to cluster.hocon. After upgrade, EMQX will continue to read local-override.conf (if it exists) as before, but you are encouraged to merge the configs to emqx.conf.

• #10164 Add CRL check support for TLS MQTT listeners.

• #10207 Improve OpenAPI (swagger) document readability. Prior to this change, there were a few summary docs which are lengthy and lack of translation, now it makes use of the more concise label field from schema i18n database instead.

• #10210 Eliminated a few harmless error level logs. Prior to this change, there might be some Mnesia callback (hook) failures occasionally occurring when stopping/restarting Mria. Now the callbacks (hooks) are unregistered prior to stop. See also Mria PR.

• #10224 Add the option to customize clusterIP in Helm chart, so that a user may set it to a fixed IP.

• #10263 Add command eval-ex for Elixir expression evaluation.

• #10278 Refactor the directory structure of all gateways.

• #10206 Support async query mode for all data bridges.

  Prior to this change, setting the query mode of a resource such as a bridge to sync would force the buffer to call the underlying connector in a synchronous way, even if it supports async calls.

• #10306 Add support for async query mode for most bridges.

  This is a follow-up change after #10206. Before this change, some bridges (Cassandra, MongoDB, MySQL, Postgres, Redis, RocketMQ, TDengine) were only allowed to be created with a sync query mode. Now async mode is also supported.

• #10318 Prior to this enhancement, only double quotes (") were allowed in rule engine SQL language's FROM clause. Now it also supports single quotes (').

• #10336 Add /rule_engine API endpoint to manage configuration of rule engine.

• #10354 More specific error messages when configure with bad max_heap_size value. Log current value and the max value when the message_queue_too_long error is thrown.

• #10358 Hide flapping_detect/conn_congestion/stats configuration. Deprecate flapping_detect.enable.

• #10359 Metrics now are not implicitly collected in places where API handlers don't make any use of them. Instead, a separate backplane RPC gathers cluster-wide metrics.

• #10373 Deprecate the trace.payload_encode configuration. Add payload_encode=[text,hidden,hex] option when creating a trace via HTTP API.

• #10381 Hide the auto_subscribe configuration items so that they can be modified later only through the HTTP API.

• #10391 Hide a large number of advanced options to simplify the configuration file.

  That includes rewrite, topic_metric, persistent_session_store, overload_protection, flapping_detect, conn_congestion, stats,auto_subscribe, broker_perf, shared_subscription_group, slow_subs, ssl_options.user_lookup_fun and some advance items in node and dashboard section, #10358, #10381, #10385.

• #10404 Change the default queue mode for buffer workers to memory_only. Before this change, the default queue mode was volatile_offload. When under high message rate pressure and when the resource is not keeping up with such rate, the buffer performance degraded a lot due to the constant disk operations.

• #10140 Integrate Cassandra into bridges as a new backend. At the current stage only support Cassandra version 3.x, not yet 4.x.

• #10143 Add RocketMQ data integration bridge.

• #10165 Support escaped special characters in InfluxDB data bridge write_syntax. This update allows to use escaped special characters in string elements in accordance with InfluxDB line protocol.

• #10211 Hide broker.broker_perf config and API documents. The two configs route_lock_type and trie_compaction are rarely used and requires a full cluster restart to take effect. They are not suitable for being exposed to users. Detailed changes can be found here: https://gist.github.com/zmstone/01ad5754b9beaeaf3f5b86d14d49a0b7/revisions.

• #10294 When configuring a MongoDB bridge, you can now use the ${field} syntax to reference fields in the message. This enables you to select the collection to insert data into dynamically.

• #10363 Implement Microsoft SQL Server bridge.

• #10573 Improved performance of Webhook bridge when using synchronous query mode. This also should improve the performance of other bridges when they are configured with no batching.

Bug Fixes

• #10145 Add field status_reason to GET /bridges/:id response in case this bridge is in status disconnected if internal health-check reports an error condition. Include this same error condition in message when creating an alarm for a failing bridge.

• #10172 Fix the incorrect regular expression in default ACL rule to allow specify username(dashboard) to subscribe $SYS/#.

• #10174 Upgrade library esockd from 5.9.4 to 5.9.6. Fix an unnecessary error level logging when a connection is closed before proxy protocol header is sent by the proxy.

• #10195 Add labels to API schemas where description contains raw HTML, which would break formatting of generated documentation otherwise.

• #10196 Use lower-case for schema summaries and descriptions to be used in menu of generated online documentation.

• #10209 Fix bug where a last will testament (LWT) message could be published when kicking out a banned client.

• #10225 Allow installing a plugin if its name matches the beginning of another (already installed) plugin name. For example: if plugin emqx_plugin_template_a is installed, it must not block installing plugin emqx_plugin_template.

• #10226 Handle validation error in /bridges API and return 400 instead of 500.

• #10242 Fixed a log data field name clash. Prior to this fix, some debug logs may report a wrong Erlang PID which may affect troubleshooting session takeover issues.

• #10257 Fixed the issue where auto_observe was not working in LwM2M Gateway.

  Before the fix, OBSERVE requests were sent without a token, causing failures that LwM2M clients could not handle.

  After the fix, LwM2M Gateway can correctly observe the resource list carried by client, furthermore, unknown resources will be ignored and printing the following warning log:

  2023-03-28T18:50:27.771123+08:00 [warning] msg: ignore_observer_resource, mfa: emqx_lwm2m_session:observe_object_list/3, line: 522, peername: 127.0.0.1:56830, clientid: testlwm2mclient, object_id: 31024, reason: no_xml_definition

• #10286 Enhance logging behaviour during boot failure. When EMQX fails to start due to corrupted configuration files, excessive logging is eliminated and no crash dump file is generated.

• #10297 Keeps eval command backward compatible with v4 by evaluating only Erlang expressions, even on Elixir node. For Elixir expressions, use eval-ex command.

• #10300 Fixed issue with Elixir builds that prevented plugins from being configured via environment variables.

• #10315 Fix crash checking limit and page parameters in /mqtt/delayed/messages API call.

• #10317 Do not expose listener level authentications before extensive verification.

• #10323 For security reasons, the value of the password field in the API examples is replaced with ******.

• #10410 Fix config check failed when gateways are configured in emqx.conf. This issue was first introduced in v5.0.22 via #10278, the boot-time config check was missing.

• #10533 Fixed an issue that could cause (otherwise harmless) noise in the logs.

  During some particularly slow synchronous calls to bridges, some late replies could be sent to connections processes that were no longer expecting a reply, and then emit an error log like:

  2023-04-19T18:24:35.350233+00:00 [error] msg: unexpected_info, mfa: emqx_channel:handle_info/2, line: 1278, peername: 172.22.0.1:36384, clientid: caribdis_bench_sub_1137967633_4788, info: {#Ref<0.408802983.1941504010.189402>,{ok,200,[{<<"cache-control">>,<<"max-age=0, ...">>}}

  Those logs are harmless, but they could flood and worry the users without need.

• #10449 Validate the ssl_options and header configurations when creating authentication http (authn_http). Prior to this, incorrect ssl configuration could result in successful creation but the entire authn being unusable.

• #10548 Fixed a race condition in the HTTP driver that would result in an error rather than a retry of the request. Related fix in the driver: emqx/ehttpc#45

• #10201 In TDengine data bridge, removed the redundant database name from the SQL template.

• #10270 ClickHouse data bridge has got a fix that makes the error message better when users click the test button in the settings dialog.

• #10324 Previously, when attempting to reconnect to a misconfigured ClickHouse bridge through the dashboard, users would not receive an error message. This issue is now resolved, and error messages will now be displayed.

• #10438 Fix some configuration item terminology errors in the DynamoDB data bridge:

  • Changed database to table
  • Changed username to aws_access_key_id
  • Changed password to aws_secret_access_key

5.0.3

Release Date: 2022-07-07

Bug fixes

• Websocket listener failed to read headers X-Forwared-For and X-Forwarded-Port 8415
• Deleted cluster_singleton from MQTT bridge config document. This config is no longer applicable in 5.0 8407
• Fix emqx/emqx:latest docker image publish to use the Erlang flavor, but not Elixir flavor 8414
• Changed the exp field in JWT auth to be optional rather than required to fix backwards compatability with 4.X releases. 8425

Enhancements

• Improve the speed of dashboard's HTTP API routing rule generation, which sometimes causes timeout 8438

5.0.2

Release Date: 2023-04-12

Enhancements

• #10022 Release installation packages for Rocky Linux 9 (compatible with Red Hat Enterprise Linux 9) and macOS 12 for Intel platform.

• #10139 Add extraVolumeMounts to EMQX Helm Chart, you can mount user's own files to EMQX instance, such as ACL rule files mentioned in #9052.

• #9893 When connecting with the flag clean_start=false, EMQX will filter out messages that published by clients banned by the blacklist feature in the session. Previously, messages sent by clients banned by the blacklist feature could still be delivered to subscribers in this case.

• #9986 Add MQTT ingress to helm charts and remove obsolete mgmt references.

• #9564 Implement Kafka Consumer Bridge, which supports consuming messages from Kafka and publishing them to MQTT topics.

• #9881 Improve error logging related to health checks for InfluxDB connections.

• #9985 Implement ClickHouse Data Bridge

• #10123 Improve the performance of /bridges API. Earlier, when the number of nodes in the cluster was large or the node was busy, the API may had a request timeout.

• #9998 Obfuscate request body in error log when using HTTP service for client authentication for security reasons.

• #10026 Metrics are now only exposed via the /bridges/:id/metrics endpoint, and no longer returned in other API operations.

• #10052 Improve startup failure logs in daemon mode.

Bug Fixes

• #10013 Fix return type structure for error case in API schema for /gateways/:name/clients.

• #10014 Ensure Monitor API /monitor(_current)/nodes/:node returns 404 instead of 400 if node does not exist.

• #10027 Allow setting node name via environment variable EMQX_NODE__NAME in Docker.

• #10050 Ensure Bridge API returns 404 status code consistently for resources that don't exist.

• #10055 The configuration parameter mqtt.max_awaiting_rel was not functional and has now been corrected.

• #10056 Fix /bridges API status code. Return 400 instead of 403 in case of removing a data bridge that is dependent on an active rule. Return 400 instead of 403 in case of calling operations (start|stop|restart) when Data-Bridging is not enabled.

• #10066 Improve error messages for /briges_probe and [/node/:node]/bridges/:id/:operation API calls to make them more readable. And set HTTP status code to 400 instead of 500.

• #10074 Check if type in PUT /authorization/sources/:type matches type given in the request body.

• #10079 Fix wrong description about shared_subscription_strategy.

• #10085 Consistently return 404 for all requests on non-existent source in /authorization/sources/:source[/*].

• #10098 Fix an issue where the MongoDB connector crashed when MongoDB authorization was configured.

• #10100 Fix channel crash for slow clients with enhanced authentication. Previously, when the client was using enhanced authentication, but the Auth message was sent slowly or the Auth message was lost, the client process would crash.

• #10107 For operations on Bridges API if bridge-id is unknown we now return 404 instead of 400.

• #10117 Fix an error occurring when a joining node doesn't have plugins that are installed on other nodes in the cluster. After this fix, the joining node will copy all the necessary plugins from other nodes.

• #10118 Fix problems related to manual joining of EMQX replicant nodes to the cluster.

• #10119 Fix crash when statsd.server is set to an empty string.

• #10124 The default heartbeat period for MongoDB has been increased to reduce the risk of too excessive logging to the MongoDB log file.

• #10130 Fix garbled config display in dashboard when the value is originally from environment variables.

• #10132 Fix some error logs generated by systemctl stop emqx command. Prior to the fix, the command was not stopping jq and os_mon applications properly.

• #10144 Fix an issue where emqx cli failed to set the Erlang cookie when the emqx directory was read-only.

• #10154 Change the default resume_interval for bridges and connectors to be the minimum of health_check_interval and request_timeout / 3 to resolve issue of request timeout.

• #10157 Fix default rate limit configuration not being applied correctly when creating a new listener.

• #10237 Ensure we return 404 status code for unknown node names in /nodes/:node[/metrics|/stats] API.

• #10251 Fix an issue where rule dependencies were not prompted when deleting an ingress-type bridge in use.

• #10313 Ensure that when the core or replicant node starting, the cluster-override.conf file is only copied from the core node.

• #10327 Don't increase “actions.failed.unknown” rule metrics counter upon receiving unrecoverable data bridge errors.

• #10095 Fix an issue where when the MySQL connector was in batch mode, clients would keep querying the server with unnecessary PREPARE statements on each batch, possibly causing server resource exhaustion. Footer

5.0.2

Release Date: 2022-07-02

Announcemnet: EMQX team has decided to stop supporting relup for opensouce edition. Going forward, it will be an enterprise only feature.

Main reason: relup requires carefully crafted upgrade instructions from ALL previous versions.

For example, 4.3 is now at 4.3.16, we have 4.3.0->4.3.16, 4.3.1->4.3.16, ... 16 such upgrade paths in total to maintain. This had been the biggest obstacle for EMQX team to act agile enought in deliverying enhancements and fixes.

Bug fixes

• Fixed a typo in bin/emqx which affects macOS release when trying to enable Erlang distribution over TLS 8398

5.0.1

Release Date: 2023-03-10

Enhancements

• #10019 Add low-level tuning settings for QUIC listeners.
• #10059 Errors returned by rule engine API are formatted in a more human-readable way rather than dumping the raw error including the stack trace.
• #9213 Add pod disruption budget to helm chart
• #9949 QUIC transport Multistreams support and QUIC TLS ca-cert support.
• #9932 Integrate TDengine into bridges as a new backend.
• #9967 New common TLS option 'hibernate_after' to reduce memory footprint per idle connection, default: 5s.

Bug Fixes

• #10009 Validate bytes param to GET /trace/:name/log to not exceed signed 32bit integer.

• #10015 To prevent errors caused by an incorrect EMQX node cookie provided from an environment variable, we have implemented a fail-fast mechanism. Previously, when an incorrect cookie was provided, the command would still attempt to ping the node, leading to the error message 'Node xxx not responding to pings'. With the new implementation, if a mismatched cookie is detected, a message will be logged to indicate that the cookie is incorrect, and the command will terminate with an error code of 1 without trying to ping the node.

• #10020 Fix bridge metrics when running in async mode with batching enabled (batch_size > 1).

• #10021 Fix error message when the target node of emqx_ctl cluster join command is not running.

• #10032 When resources on some nodes in the cluster are still in the 'initializing/connecting' state, the bridges/ API will crash due to missing Metrics information for those resources. This fix will ignore resources that do not have Metrics information.

• #10041 For InfluxDB bridge, added integer value placeholder annotation hint to write_syntax documentation. Also supported setting a constant value for the timestamp field.

• #10042 Improve behavior of the replicant nodes when the core cluster becomes partitioned (for example when a core node leaves the cluster). Previously, the replicant nodes were unable to rebalance connections to the core nodes, until the core cluster became whole again. This was indicated by the error messages: [error] line: 182, mfa: mria_lb:list_core_nodes/1, msg: mria_lb_core_discovery divergent cluster.

• #10054 Fix the problem that the obfuscated password is used when using the /bridges_probe API to test the connection in Data-Bridge.

• #10058 Deprecate unused QUIC TLS options. Only following TLS options are kept for the QUIC listeners:

  • cacertfile
  • certfile
  • keyfile
  • verify

• #10076 Fix webhook bridge error handling: connection timeout should be a retriable error. Prior to this fix, connection timeout was classified as unrecoverable error and led to the request being dropped.

• #10078 Fix an issue that invalid QUIC listener setting could cause a segfault.

• #10084 Fix the problem when joining core nodes running different EMQX versions into a cluster.

  Mria PR

• #10086 Upgrade HTTP client ehttpc to 0.4.7. Prior to this upgrade, HTTP clients for authentication, authorization, and webhook may crash if body is empty but content-type HTTP header is set. For more details see ehttpc PR#44.

• #9939 Allow 'emqx ctl cluster' command to be issued before Mnesia starts. Prior to this change, EMQX replicant could not use manual discovery strategy. Now it's possible to join cluster using 'manual' strategy.

• #9958 Fix the error code and error message returned by the clients API when the Client ID does not exist.

• #9961 Avoid parsing config files for node name and cookie when executing non-boot commands in bin/emqx.

• #9974 Report memory usage to statsd and Prometheus using the same data source as dashboard. Prior to this fix, the memory usage data source was collected from an outdated source which did not work well in containers.

• #9997 Fix Swagger API schema generation. deprecated metadata field is now always boolean, as Swagger specification suggests.

• #10007 Change Kafka bridge's config memory_overload_protection default value from true to false. EMQX logs case when messages get dropped due to overload protection, and this is also reflected in counters. However, since there is by default no alerting based on the logs and counters, setting it to true may cause messages being dropped without notice. At the time being, the better option is to let sysadmin set it explicitly so they are fully aware of the benefits and risks.

• #10087 Use default template ${timestamp} if the timestamp config is empty (undefined) when inserting data in InfluxDB. Prior to this change, InfluxDB bridge inserted a wrong timestamp when template is not provided.

5.0.1

Release Date: 2022-07-01

Enhancements

• Removed management API auth for prometheus scraping endpoint /api/v5/prometheus/stats 8299
• Added more TCP options for exhook (gRPC) connections. 8317
• HTTP Servers used for authentication and authorization will now indicate the result via the response body. 8374 8377
• Bulk subscribe/unsubscribe APIs 8356
• Added exclusive subscription 8315
• Provide authentication counter metrics 8352 8375
• Do not allow admin user self-deletion 8286
• After restart, ensure to copy cluster-override.conf from the clustered node which has the greatest tnxid. 8333

Bug fixes

• A bug fix ported from 4.x: allow deleting subscriptions from client.subscribe hookpoint callback result. 8304 8347
• Fixed Erlang distribution over TLS 8309
• Made possible to override authentication configs from environment variables 8323
• Made authentication passwords in Mnesia database backward compatible to 4.x, so we can support data migration better. 8351
• Fix plugins upload for rpm/deb installations 8379
• Sync data/authz/acl.conf and data/certs from clustered nodes after a new node joins the cluster 8369
• Ensure auto-retry of failed resources 8371
• Fix the issue that the count of packets.connack.auth_error is inaccurate when the client uses a protocol version below MQTT v5.0 to access 8178

Others

• Rate limiter interface is hidden so far, it's subject to a UX redesign.
• QUIC library upgraded to 0.0.14.
• Now the default packages will be released withot otp version number in the package name.
• Renamed config exmpale file name in etc dir.

5.0.0

Release Date: 2022-06-17

Horizontal Scalability

An extension to the Mnesia database was introduced, named 'Mria'.

EMQX nodes can continue to form a cluster (named 'core') exactly as in previous versions.

In version 5, Mria allows more nodes to join the cluster as 'replicants'.

The difference is, replicant nodes apply changes in the routing table (and configurations etc.) asynchronously, and the replicant nodes can be scaled up and down without affecting the data consistency.

This reduces the networking overheads of forming a full-mesh cluster.

In our most recent test, we managed to achieve 100 million MQTT clients connecting to a 23 nodes Mria cluster.

It's a 3 times larger cluster size compared to a typical version 4 cluster,

and 10 times more capacity comparing to our previous 10 million achievement

Reliability

With the help of the Mria clustering topology, the risk of suffering a split-view of the cluster due to network partitioning, and improves cluster stability in general.

• Only core nodes are involved in transactional writes
• Data is fully replicated to replicant nodes so they can read from local copy (i.e., from RAM)
• While core nodes are fully functional MQTT broker, one can also opt to dedicate them as pure data base nodes

The new Dashboard

Version 5 comes with a fresh new design for the dashboard. Powered by vue.js at the frontend, and OpenAPI at the backend, it provides the most easy-to-use MQTT broker management UI.

Key improvements compared to version 4:

• Access control (authentication and authorization) management
• Improved rule engine and action management UI
• Visual aid view to display data flow e.g., from MQTT topics to destination data sink
• Online config update
• Gateway/Extension management
• More diagnosis tools, such as slow subscriptions and online tracing

Typesafe Config

Starting from version 5, EMQX will use HOCON for the configuration syntax.

With a schema on top, the configs are now type-checked.

HOCON supports NGINX-like configuration layouts, and provides a native array syntax.

But don't worry, the old cuttlefish syntax is still supported if you prefer to keep everything flat like in version 4.

The EMQX team have put huge effort into providing sane-default values, as a result, the default configuration file is now less than 100 lines including comments.

Provided side-by-side, there is a full-blown example configuration (generated from the schema) to help us quickly find examples to copy from.

Operability

In version 5, the interface descriptions for the Dashboard-enabled management APIs follow the OpenAPI specification 3.0. Behind the scene, the specification is actually generated from the configuration schema, the same schema which type-checks the files.

Now we have a single source of schema, which is used to guard both the configuration file and HTTP interface, also to generate config and API documents.

Another nice thing about Swagger, is that it comes with the Swagger-UI in which we can click a "try it out" button to test the API directly from the web browser, or copy-paste the curl command example to a console to test it out.

Another major change in operability compared to version 4 is that the changes are persisted back on disk as configuration file (in HOCON syntax).

The immediate benefit from unifying the config file interface and HTTP API interface is the hot-configuration.

Hot-configuration was a feature only available in EMQX Enterprise edition, it allows us to change a lot of configurations from Dashboard or through the APIs at runtime without having to restart the service.

In enterprise edition version 4, the hot-values were stored in the database, but not in config files, which was not very convenient for users who want to change the overridden values from the config file interface.

Observability

The dashboard comes with more detailed monitoring metrics. We can now view up to 7 days of historical metrics data on Dashboard, and integrate with Prometheus with one click. Add log tracking and slow subscription diagnostic tools to effectively improve the user experience when troubleshooting and diagnosing abnormal client behaviors.

Another enhancement in observability is the introduction of structured logging, now most of the logs emitted from EMQX have a msg field. The text of this filed is an underscore-separated words, to make it more search-friendly for humans but also helps the log indexing tools to index the logs.

Data Integration (The Old Rule Engine)

As you may notice from the dashboard, the old 'Rule Engine' has been renamed to 'Data integration'.

It includes two major functionalities: rules and data bridges.

Rule is a data processing language in SQL syntax (in addition a jq language support), which helps to filter and transform IoT messages.

Data bridge provides the ability to ingest data into EMQX or export data outside of EMQX.

Through the dashboard, users can now clearly see how IoT data is processed with rules and how the data flows from/to external data services.

Another important change: now rules and actions are configs (in config files), but not data in Mnesia (the built-in database, which is quite opaque for non-Erlang developers), meaning we'll be able to configure rules with various deployment/orchestration tools a lot easier.

Access Control

Now we can manage authentication and authorization from the dashboard without having to restart the broker.

Previously provided as plugins, authentication and authorization (ACL) configurations are scattered in different files, and changes will often require a restart of the broker to take effect. In version 5, all the commonly used security configurations are grouped together. There is also a user management UI provided in the dashboard, allowing you to manage users and access rules on the fly.

We can even configure different authentication rules per listener.

Gateway

The gateway is re-implemented with a unified design language, providing independent management interfaces and security authentication capabilities for various protocols with different client attributes and life cycles, allowing users to access in a more native way. Since each gateway can be configured with its own independent authentication, authentication information of different gateway devices can now be isolated from each other to meet the needs of more scenarios.

MQTT over QUIC

QUIC RFC 9000, the next generation transport layer for the internet, brings amazing opportunities (and challenges too) to IoT.

EMQX team cannot wait to start experimenting MQTT on top of it.

Coming with the 5.0 release, one can configure a QUIC listener to experiment with MQTT over QUIC.

Extension

EMQX 5.0 keeps and to a certain extent enhances the ability of plugin extensions. The plugins can now be compiled, distributed as a standalone package, and then uploaded on the Dashboard to complete the installation for the entire cluster, so they do not have to repeat the steps for each node.

Hello Elixir

EMQX 5 is still a rebar3 project, but it now compiles with mix, if you know Elixir, you know what it means. Happy hacking.