Repository navigation
fix(deployment): simplify probes command to reduce false positives - #135
Merged
Merged
Conversation
polarathene
reviewed
Sep 7, 2024
The double-grep behavior led to the command unintentionally hanging in some cases, prompting k8s to continually restart the pod without valid reason.
note: there is optional work remaining to conditionally add more services to liveness probe based on what the user has chosen to enable (or things that may perhaps always be runnig).
fix: update test snapshots
Member
Author
|
Note the PR now has been adapted for the dms-healthcheck script, but until it lands the chart-testing steps will fail. Will revisit once a new release is cut to target against. |
Collaborator
|
This looks good to me. We can merge once a new version of docker-mailserver is released. @polarathene any thoughts on that? |
Member
Release is blocked on me finding time to review the Debian 13 + Dovecot 2.4 upgrade PR 😓 Unfortunately DMS is low on maintainers, and PRs like that one aren't as breezy to review confidently 😅 I'll try to fit some time towards resuming my review this week if I can. |
cfis
added a commit
to cfis/docker-mailserver-helm
that referenced
this pull request
Oct 27, 2026
The 6.0.0 version bump moves to the probes change (docker-mailserver#135), which depends on docker-mailserver 16, so both are released together. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
kube-score 1.20 checks for identical liveness and readiness probes separately from pod-probes, which was already ignored. dms-healthcheck is the only health signal docker-mailserver provides, so the probes intentionally use it for both, with a startup probe covering slow starts. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Collaborator
|
Thanks @radicand, and sorry this sat so long! I just updated docker-mailserver to version 16 so I have updated your PR and merged it. Thanks! |
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
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.
NOTE This PR requires the chart to first be updated for DMS v16.
In some (rare - one documented in #134) instances, the deployment liveness/readiness probes may time out while running. Investigation into this seems to be that the
supervisorctl statuscommand returned the right information, but somewhere in the double grep chaining the command could hang.The good news is there is no need for the double grepping.
supervisorctl statuswill return an exit code of0when all services provided as arguments are in theRUNNINGstate, which is what we're intending to check anyway. If any services are not running,supervisorctl statusreturns a non-zero exit code which k8s will interpret as a probe failure.This cleaner method is actually more reliable regardless, as it could be the case in the previous implementation that the probe would not fail if one of the key services was not running, but any one of the specified services was.
Finally, note that in this pull request I only included the baseline services that all installations use - there is an opportunity to conditionally include more services in the check based on what the user has chosen to enable in their
values.yaml.