> 8. Two Follower SCMs shutdown due to state machine failure after SCM leader transfer. It's highly suspected to be related with raft log reapply.
Is there a JIRA? If not, please file one. We can debug it together. Tsz-Wo On Fri, Apr 25, 2025 at 1:25 AM Sammi Chen <sammic...@apache.org> wrote: > Attenders: Pony, Hao, Minyu, Lansong, Peter, Chung en, Wei-Chiu, Sammi > > 1. Disk Balancer - development ongoing. > 2. Object TTL - first patch under review. > 3. Short-circuit read - system test ongoing. > 4. 2.0.0 release - test is close to finish, most likely will be released > this week. > 5. 2.0.1 release - under discussion. > 6. A new proposal to support read from follower without going through ratis > consensus, for use cases who can tolerate potential stale data or no big > chance to get stale data. > 7. S3 compatibility - discussed S3 event notification investigation. > 8. Two Follower SCMs shutdown due to state machine failure after SCM leader > transfer. It's highly suspected to be related with raft log reapply. > > > 2025-04-25 11:09:34,000 > [f82d66d9e701-4134-9017-100ed7a75d18@group6E13D37CF212-StateMachineUpdater > ] > ERROR org.apache.ratis.statemachine.StateMachine: Terminating with exit > status 1: Invalid event: CLOSE at OPEN state. > > > > org.apache.hadoop.ozone.common.statemachine.InvalidStateTransitionException: > Invalid event: CLOSE at OPEN state. >
