[Serve] Let the controller look up the head node and fix flaky standalone3 healthz test #36878
Conversation
This file contains 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
Signed-off-by: Gene Su <e870252314@gmail.com>
edoakes
reviewed
Jun 27, 2023
Signed-off-by: Gene Su <e870252314@gmail.com>
Signed-off-by: Gene Su <e870252314@gmail.com>
…e look up in http state Signed-off-by: Gene Su <e870252314@gmail.com>
edoakes
reviewed
Jun 28, 2023
Signed-off-by: Gene Su <e870252314@gmail.com>
Signed-off-by: Gene Su <e870252314@gmail.com>
sihanwang41
approved these changes
Jun 28, 2023
GeneDer
changed the title
|
Sorry, i misread the code. the controller is already stick with head node. We just potentially pass wrong head_node_id to the controller. (this pr can fix it)
|
Signed-off-by: Gene Su <e870252314@gmail.com>
GeneDer
changed the title
|
@GeneDer |
Signed-off-by: Gene Su <e870252314@gmail.com>
Signed-off-by: Gene Su <e870252314@gmail.com>
Signed-off-by: Gene Su <e870252314@gmail.com>
Signed-off-by: Gene Su <e870252314@gmail.com>
Signed-off-by: Gene Su <e870252314@gmail.com>
Signed-off-by: Gene Su <e870252314@gmail.com>
Signed-off-by: Gene Su <e870252314@gmail.com>
Signed-off-by: Gene Su <e870252314@gmail.com>
Signed-off-by: Gene Su <e870252314@gmail.com>
Signed-off-by: Gene Su <e870252314@gmail.com>
…orker.client() Signed-off-by: Gene Su <e870252314@gmail.com>
Signed-off-by: Gene Su <e870252314@gmail.com>
Signed-off-by: Gene Su <e870252314@gmail.com>
Signed-off-by: Gene Su <e870252314@gmail.com>
Signed-off-by: Gene Su <e870252314@gmail.com>
8 tasks
arvind-chandra
pushed a commit
to lmco/ray
that referenced
this pull request
Aug 31, 2023
…lone3 healthz test (ray-project#36878) - Make sure we are using wait_for_condition in the test (could take time to broadcast). - Remove head_node_id from controller init args and instead fetch it in the controller init. Also remove it from serve_start in _private/api.py. - Add an assertion to check that the controller actually runs on the head node (use ray.nodes() and look for head node resource). - Filter Nones from the active node set in deployment_state. Add a unit test for this, it should never return None. Signed-off-by: e428265 <arvind.chandramouli@lmco.com>
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.
Why are these changes needed?
wait_for_conditionin the test (could taketime to broadcast).
head_node_idfrom controller init args and instead fetch itin the controller init. Also remove it from
serve_startin_private/api.py.head node (use ray.nodes() and look for head node resource).
Nones from the active node set in deployment_state. Add aunit test for this, it should never return
None.Related issue number
Closes: #37001
Fixes flaky test: https://buildkite.com/ray-project/oss-ci-build-branch/builds/4687#0188fdc4-2bd2-4e43-828c-d92f6d48e89d
Checks
git commit -s) in this PR.scripts/format.shto lint the changes in this PR.method in Tune, I've added it in
doc/source/tune/api/under thecorresponding
.rstfile.