Repository navigation
fix(docker): probe local ClickHouse health over IPv4 - #962
Conversation
The dev compose healthcheck ran wget against localhost:8123, which resolves to [::1] inside the container while ClickHouse listens on IPv4, so the container reported unhealthy forever (84,642 consecutive failures locally) regardless of the server's state. docker-compose.selfhost.yml already probes 127.0.0.1 since #804.
|
The latest updates on your projects. Learn more about Vercel for GitHub.
|
|
The latest updates on your projects. Learn more about Unkey Deploy
|
|
Important Review skippedAuto reviews are disabled on this repository. Please check the settings in the CodeRabbit UI or the ⚙️ Run configurationConfiguration used: Repository UI Review profile: ASSERTIVE Plan: Advanced Run ID: You can disable this status message by setting the Use the checkbox below for a quick retry:
Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out. Comment |
|
Summary
The dev
docker-compose.yamlhealthcheck ranwget http://localhost:8123/ping. Inside the containerlocalhostresolves to[::1], but ClickHouse listens on IPv4, so the probe is refused every time and the container reportsunhealthyforever (84,642 consecutive failures on one machine) regardless of whether the server is fine. That hides real outages: a ClickHouse that is actually hung looks identical to a healthy one.docker-compose.selfhost.ymlalready probes127.0.0.1since #804; this applies the same fix to the dev compose file.Test plan
wget --spider http://127.0.0.1:8123/pingsucceeds; thelocalhostform is refuseddocker compose -f docker-compose.yaml config -qSummary by cubic
Fixes the dev ClickHouse healthcheck so it probes IPv4 instead of
localhost, which resolved to[::1]inside the container and never matched ClickHouse's IPv4 listener. The container now correctly reports healthy;docker-compose.selfhost.ymlalready used127.0.0.1.Written for commit 0aec854. Summary will update on new commits.