LexoRank rebalancing job may stuck in some cases. You don't see any progress for many hours (days).
You can see status at Lexorank management page (see LexoRank Management)
- Example of REST call /rest/greenhopper/1.0/lexorank/balance output:
- Note progress stuck at 98%.
- Please note that
JSWSERVER-15703introduces the back-off mechanism for JIRA DC, so this is expected that Lexorank Rebancing pauses under high load:
Whenever we detect current node is behind any other node in the cluster by 30 seconds or any other node in the cluster is behind current node by 30 seconds, LexoRank balancing will back-off.
Lexorank Rebalancing job is running and has progress.
Lexorank Rebalancing job is stuck and has no progress.
User impact: Lexorank Rebalancing fixes issues with "broken" Rank, if current progress did not fix those broken issues so far, user will not able to rank them.
- Thread dump generated at that time will show the lexorank-executor-thread-0 is WAITING for a long time:
- Enabling DEBUG logging Lexorank balance (com.atlassian.greenhopper.service.lexorank.balance) doesn't show any progress
- Expected logging:
- See KB for more details regarding rebalancing: Troubleshooting New Ranking System Issues