Repository navigation
refactor: the observability services run in every stack, so local and production start the same services - #456
Merged
Conversation
… production start the same services Production's .env set COMPOSE_PROFILES=observability and local stacks ran without Grafana, Loki, OpenSearch, VictoriaMetrics, Jaeger and the collectors unless asked. The profile goes, the deploy no longer writes COMPOSE_PROFILES, and Kafdrop stays under the debug profile. Compose leaves profiles out of its config hash, so the release recreates nothing.
HardMax71
enabled auto-merge (squash)
October 23, 2026 18:29
|
Important
This repository does not receive automatic reviews because it has fewer than 10 stars. ⚙️ Run configuration
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 |
github-actions
Bot
added
documentation
|
|
Codecov Report✅ All modified and coverable lines are covered by tests.
Flags with carried forward coverage won't be shown. Click here to find out more.
🚀 New features to boost your workflow:
|
HardMax71
added a commit
that referenced
this pull request
Oct 23, 2026
…the observability services No E2E test uses Grafana, Loki, OpenSearch, VictoriaMetrics, Jaeger or the collectors. Since #456 the E2E jobs pulled their 4 GB of images too: the pull grew from 22 s to 49 s and the stack waited about 25 s on it, plus 13 s more startup.
6 of 17 tasks
HardMax71
added a commit
that referenced
this pull request
Oct 23, 2026
… and E2E skips Playwright's apt step and the observability images (#457) * fix(ci): each uv job keeps its own cache, Playwright skips its apt step, and grype's database is cached All backend jobs shared one setup-uv key, so the first job to finish (bandit only) saved a 3 MB cache and the others downloaded CPython and every wheel on each run; each job now saves its own cache with the managed Python. The Playwright browser cache follows the browser builds in playwright-core's browsers.json instead of the whole package-lock.json, and the install-deps step goes: on the runner image it only adds fonts no test uses, and its apt download stalled two runs for more than 6 minutes. The SBOM scan restores grype's vulnerability database from a cache keyed on the checksum of the newest database, instead of spending about 80 s downloading and unpacking it in each job. * fix(ci): E2E pulls and starts only the app and its dependencies, not the observability services No E2E test uses Grafana, Loki, OpenSearch, VictoriaMetrics, Jaeger or the collectors. Since #456 the E2E jobs pulled their 4 GB of images too: the pull grew from 22 s to 49 s and the stack waited about 25 s on it, plus 13 s more startup.
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.
Summary
Production's .env set
COMPOSE_PROFILES=observability, so production ran Grafana, Loki, OpenSearch, VictoriaMetrics, Jaeger and both collectors while a local stack ran without them unless asked. That was a behavior difference between local and production deploys, where only keys and credentials should differ. Follows #455.Changes
The eight observability services lose
profiles: ["observability"], sodocker compose up -dstarts them everywhere, and the deploy stops writingCOMPOSE_PROFILES. Compose leavesprofilesout of its config hash (hash.go), so the release recreates nothing on production. Kafdrop stays under thedebugprofile, since production never runs it. CI's E2E stacks now start the observability services too. The docs drop the profile everywhere.Type of Change
Test Plan
cd backend && uv run pytest tests/unit)cd frontend && npm run test)docker compose exec -T backend sh -c 'uv sync --group test --no-dev --frozen --no-install-project --no-build && pytest tests/e2e')cd frontend && npm run test:e2e)uv run python -m scripts.export_openapiin backend/, thennpm run generate:apiin frontend/), never edited by handValidation:
docker compose config --serviceslists the observability services with no profile set and Kafdrop only with--profile debug, andmkdocs build --strictpasses.Checklist
pre-commit run --all-filesand all checks passSummary by cubic
Makes the observability services part of every compose stack, so local and production now start the same services: Grafana, Jaeger, Loki, OpenSearch, VictoriaMetrics, and both OTel collectors come up with
docker compose up -d.Changes
observabilityprofile from all eight services; production's deploy no longer writesCOMPOSE_PROFILES=observability.profilesout of its config hash, so the production release recreates no containers.debugprofile, and CI's E2E stacks now start the observability services.Written for commit c317bc4. Summary will update on new commits.