Skip to content

[FIX] Fix flaky BookieAutoRecoveryTest#testOpenLedgers timeout#4743

Merged
merlimat merged 1 commit intoapache:masterfrom
void-ptr974:fix_flaky_test
Apr 22, 2026
Merged

[FIX] Fix flaky BookieAutoRecoveryTest#testOpenLedgers timeout#4743
merlimat merged 1 commit intoapache:masterfrom
void-ptr974:fix_flaky_test

Conversation

@void-ptr974
Copy link
Copy Markdown
Contributor

When the killed bookie happens to be the Auditor leader, the test must wait for the ZK session timeout (default 10s) before the ephemeral node disappears, then wait for a new Auditor
leader election, metadata scan, and underreplicated ledger publishing. In resource-constrained CI environments, this chain can exceed the 60-second await timeout.

Two changes:

  • setZkTimeout(4000) — reduces ZK session timeout so the ephemeral node disappears faster
  • await 60s → 90s — provides more headroom for slow CI environments

When the killed bookie happens to be the Auditor leader, the test must wait for the ZK session timeout (default 10s) before the ephemeral node disappears, then wait for a new Auditor
   leader election, metadata scan, and underreplicated ledger publishing. In resource-constrained CI environments, this chain can exceed the 60-second await timeout.

  Two changes:
  - setZkTimeout(4000) — reduces ZK session timeout so the ephemeral node disappears faster
  - await 60s → 90s — provides more headroom for slow CI environments
@void-ptr974
Copy link
Copy Markdown
Contributor Author

@merlimat merlimat merged commit b5ce0e5 into apache:master Apr 22, 2026
20 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants