cockroachdb / cockroachdb/cockroach
multitenant: toggling between none and shared service repeatedly sometimes hangs
- Dominant language
- Go
- Stars
- 32.5k
- Forks
- 4.1k
- PR merge metrics
- PR metrics pending
Description
In #119636 we observed some tests would hang for 30-40s trying to start/restart service for tenants if that tenant had previously been started. There are some indications this could be due to interactions of bootstrap upgrade jobs and/or draining but further investigation is required. Removing `WITH SERVICE NONE` from the initial tenant creation at the top of `TestRevertTenantToTimestamp` was one of the observed cases of this issue.
Jira issue: CRDB-36292
Contributor guide
Research direction
Start by reproducing the hang in TestRevertTenantToTimestamp, including the initial tenant creation with and without WITH SERVICE NONE. Investigate the reported interactions between bootstrap upgrade jobs and draining, using issue #119636 and Jira CRDB-36292 for context. Done means identifying and fixing the cause so repeatedly toggling tenants between none and shared service does not hang.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- go
- Domain
- databases, distributed-systems
- Issue type
- Bug
- Difficulty
- 4/5
- Estimated time
- 3-5 days
- Activity status
- Stale
- Clarity
- Needs clarification
- Newbie friendliness
- 25/100