Known Issues in EMQX 6.1

6.1.1

Since version	Issue	Workaround	Status
5.8.0	Core-replicant cluster startup may not complete when Durable Storage is enabled When initializing a cluster with core and replicant nodes, Durable Storage (DS) metadata may become inconsistent across nodes, causing partial unavailability of related features: Durable Sessions, Queues, or Streams. This can affect both new clusters and clusters upgraded from EMQX 5.x, especially when n_sites > 1. It is typically indicated by continuous warnings such as: msg: dsrepl_optimistic_leader_fail, reason: {init_failed,{error,recoverable,{fsm_needs_upgrade,0}}}	Stop and restart each replicant node one at a time.