Uh oh!
There was an error while loading. Please reload this page.
fix(#1143): add readiness/liveness/startup probes to the requests-proxy Deployment - #641
Merged
Merged
Conversation
The requests-proxy Deployment declared ports/securityContext/resources/env but no probes, so the pod was marked Ready the instant the container process started -- before gunicorn had bound :8888. `tb doctor`'s "Service Bus egress (requests-proxy) ready" check reads ReadyReplicas, so it showed a false green even when the proxy could not relay. Add tcpSocket startup/liveness/readiness probes on the gunicorn listen port (8888), mirroring mysql-deployment.yaml's three-probe structure. tcpSocket (not httpGet) because the proxy application lives in the separate client-runtime repo and exposes no chart-visible HTTP health path; a probe against an unknown path would 404 and wedge the pod, whereas a successful TCP accept is the honest "gunicorn is serving" signal the doctor check needs. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
saadqbal
previously approved these changes
Aug 7, 2026
saadqbal
left a comment
Contributor
There was a problem hiding this comment.
Nice, tight fix — probe timings mirror mysql-deployment, tcpSocket is the right call given there's no chart-visible HTTP path, and version/appVersion move in lockstep as usual. One non-blocking note inline about the residual false-green. Approving.
Uh oh!
There was an error while loading. Please reload this page.
aptracebloc
previously approved these changes
Aug 7, 2026
aptracebloc
left a comment
Contributor
There was a problem hiding this comment.
Looks good and I adhere to Asad's non-blocking comment.
tcpSocket:8888 confirms the gunicorn master has bound the listen socket — a strict improvement over no probes that catches a container that never starts — but it does not fully close the false-green: the master binds before the worker finishes importing create_app() and keeps the socket bound while a worker crash-loops, so a proxy that can't relay can still read Ready. A real /healthz in client-runtime would be needed to close it fully. Probes themselves unchanged. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
saadqbal
approved these changes
Aug 7, 2026
Uh oh!
There was an error while loading. Please reload this page.
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 freeto join this conversation on GitHub.
Already have an account?
Sign in to comment
4 participants
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.
Closes tracebloc/backend#1143.
Problem
The requests-proxy Deployment has no readiness/liveness/startup probes, so the pod counts as Ready the moment the container starts — before gunicorn is actually accepting connections. The doctor's
ReadyReplicascheck goes green while the proxy can't yet serve (false-green).Fix
Add
startupProbe/livenessProbe/readinessProbeastcpSocketon the gunicorn port (8888), mirroring the three-probe structure ofmysql-deployment.yaml.Why tcpSocket, not httpGet: the requests-proxy application lives in the
client-runtimerepo and exposes no chart-visible HTTP health path — anhttpGetto an unknown path would 404 and permanently wedge the pod. A successful TCP accept on 8888 is the honest "gunicorn is serving" signal and is exactly what the doctor's ReadyReplicas check needs.Files
client/templates/requests-proxy-deployment.yaml— three tcpSocket probes on port 8888client/Chart.yaml— version + appVersion → 1.9.26 (chart-content change)Validation
helm lintclean;helm templaterenders the probes as intended.🤖 Generated with Claude Code
Note
Low Risk
Helm-only change to probe timing on a single-replica sidecar; no auth or data-path logic, with documented limits of tcpSocket vs full app health.
Overview
Fixes false-green readiness for the Service Bus egress proxy:
tb doctorwas treating the pod as ready as soon as the container started, before gunicorn listened on 8888.The requests-proxy Deployment now has startup, liveness, and readiness probes as
tcpSocketon port 8888, aligned with the three-probe pattern used elsewhere in the chart (e.g. MySQL). HTTP probes were avoided because the proxy app has no chart-visible health URL; a badhttpGetcould 404 and block readiness permanently.Chart version is bumped to 1.9.26. Inline comments note this is a partial fix: TCP success only means the gunicorn master bound the socket, not that
create_app()and the worker can actually relay traffic—a future/healthzin client-runtime would be needed for that.Reviewed by Cursor Bugbot for commit 9634e74. Bugbot is set up for automated code reviews on this repo. Configure here.