Widen tenant-3 health windows for slow first boots - #24
Merged
Conversation
The 20s x 3 (~60s) TCP failure budget restarted the VM before Elasticsearch opened its transport port on a cold 1-vCPU boot, so the tiebreaker never joined the AWS tenant-3 cluster. Widen to 30s x 5, matching the budget tenant-2 kibana needed. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
artemnikitin
changed the title
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
On AWS the
tenant-3-elasticsearch-voting-onlyVM never joined the cluster: its cold first boot on 1 vCPU takes longer than the 20s × 3 (~60s) TCP health failure budget, so the agent restarted it just before the transport port opened, in a loop._cat/nodesshowed only the two data nodes — the tiebreaker (the point of the HA topology) was missing, while the same spec converged on GCP where boots ran faster.Same failure mode and same fix as tenant-2 Kibana in #22: widen the failure budget past the observed cold-start time (30s × 5 ≈ 150s+). Applied to all three tenant-3 services — the data nodes made it inside ~60s this time on 4 vCPUs, but the margin is thin and a slow cold boot would put them in the same loop; keeping the budget consistent across the cluster avoids that.
Validation
configcheck --require-remote-routing: OKservice exceeded failure threshold, restarting, repeatedly)🤖 Generated with Claude Code