Add per-LitAPI health check endpoints#712
Open
discobot wants to merge 1 commit into
Open
Conversation
Each LitAPI now exposes its own health check at {api_path}{healthcheck_path}, e.g. /predict/health, reporting worker readiness and the user-defined health() hook for that API only. The global healthcheck endpoint reuses the same per-API helper and aggregates across all APIs, so its behavior is unchanged. Two existing tests used a synthetic worker status key format the server never produces; they now use the real {endpoint}_{worker_id} format. Closes Lightning-AI#605.
discobot
requested review from
Abdul-0x4A,
andyland,
aniketmaurya,
dmitsf,
ethanwharris,
justusschock,
k223kim and
tchaton
as code owners
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
What does this PR do?
Fixes #605.
Each LitAPI now exposes its own health check at
{api_path}{healthcheck_path}— e.g./api1/health,/api2/health(and/api1/my_server/healthwith a customhealthcheck_path). The per-API check reports worker readiness for that API's workers only, plus its user-definedhealth()hook, returning the same200 "ok"/503 "not ready"contract as the global endpoint. Auth matches the API's predict route (setup_auth(lit_api)).The per-API logic lives in a single helper,
LitServer._check_lit_api_health, and the global healthcheck endpoint now aggregates it across all APIs, so the two can't drift. Worker keys are matched by exact{endpoint}_{worker_id}parsing rather than a prefix check so an API namedpredictdoesn't pick uppredict_v2's workers. Global endpoint behavior is unchanged, including 503 on emptyworkers_setup_statusand calling thehealth()hook even before workers are ready.Two existing tests stubbed
workers_setup_statuswith a synthetic"worker-0"key, a formatlaunch_inference_workernever produces; they now use the real"predict_0"format. New tests cover a two-API server (healthy API 200, unhealthy API 503, global 503) and the customhealthcheck_path+ customapi_pathcombination, parametrized over zmq like the neighboring health tests.Before submitting
As a user running multiple LitAPIs behind one server, I can now ask "is this specific API healthy?" by hitting
/api1/healthinstead of only the aggregate/health, which masks which API is degraded. This makes per-API liveness/readiness probes (Kubernetes, load balancers) and targeted alerting possible without custom workaround endpoints.PR review
Anyone in the community is free to review the PR once the tests have passed.
If we didn't discuss your PR in GitHub issues there's a high chance it will not be merged.
Did you have fun?
Yes 🙃
🤖 Implemented with the help of Claude Code.