runtime/lava: add a liveness probe and fail fast on connect - #3208
Merged
Conversation
Every device and queue query the scheduler makes against a lab that has gone away blocks for the full 30s request timeout, once per job and per platform. There is no way to ask "is this lab up?" without paying that cost on an endpoint that also does real work. Add LAVA.is_alive(), which polls /system/version/. lava-server answers it from a constant with no database access (SystemViewSet.version in lava_rest_app/v02/views.py) and the body is ~20 bytes, against ~16-78kB and a full device table scan for /devices/, so a caller can poll it regularly without adding any measurable load to the lab. Any HTTP answer counts as reachable, including the refusals: labs that return 401 or 403 to an anonymous or unprivileged request are up, and treating them as down would stop scheduling against them entirely. Only a transport failure or a 5xx is reported as unreachable. Also split the request timeout into a short connect timeout and the existing response timeout, so an unreachable host fails in seconds rather than after 30, and give Runtime a default is_alive() so callers can probe any runtime. Signed-off-by: Denys Fedoryshchenko <denys.f@collabora.com>
nuclearcat
force-pushed
the
lava-liveness
branch
from
September 8, 2026 16:56
38fc96f to
bbc6c62
Compare
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Every device and queue query the scheduler makes against a lab that has gone away blocks for the full 30s request timeout, once per job and per platform. There is no way to ask "is this lab up?" without paying that cost on an endpoint that also does real work.
Add LAVA.is_alive(), which polls /system/version/. lava-server answers it from a constant with no database access (SystemViewSet.version in lava_rest_app/v02/views.py) and the body is ~20 bytes, against ~16-78kB and a full device table scan for /devices/, so a caller can poll it regularly without adding any measurable load to the lab.
Any HTTP answer counts as reachable, including the refusals: labs that return 401 or 403 to an anonymous or unprivileged request are up, and treating them as down would stop scheduling against them entirely. Only a transport failure or a 5xx is reported as unreachable.
Also split the request timeout into a short connect timeout and the existing response timeout, so an unreachable host fails in seconds rather than after 30, and give Runtime a default is_alive() so callers can probe any runtime.