-
Notifications
You must be signed in to change notification settings - Fork 5.9k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
[release/air] Fix release test timeout for tune_scalability_network_overhead
#36360
Merged
krfricke
merged 5 commits into
ray-project:master
from
justinvyu:release/tune/test_network_overhead
Jun 16, 2023
Merged
[release/air] Fix release test timeout for tune_scalability_network_overhead
#36360
krfricke
merged 5 commits into
ray-project:master
from
justinvyu:release/tune/test_network_overhead
Jun 16, 2023
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: Justin Yu <justinvyu@anyscale.com>
Signed-off-by: Justin Yu <justinvyu@anyscale.com>
Signed-off-by: Justin Yu <justinvyu@anyscale.com>
Signed-off-by: Justin Yu <justinvyu@anyscale.com>
Release tests kicked off here: https://buildkite.com/ray-project/release-tests-pr/builds/42034 |
justinvyu
added
the
tests-ok
The tagger certifies test failures are unrelated and assumes personal liability.
label
Jun 15, 2023
krfricke
approved these changes
Jun 16, 2023
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Thanks!
8 tasks
arvind-chandra
pushed a commit
to lmco/ray
that referenced
this pull request
Aug 31, 2023
…overhead` (ray-project#36360) There are 2 versions of the test: a smoke test version with 20 nodes/trials and a full version with 100 nodes/trials. The smoke test version timed out recently. The test setups is slightly different: - Smoke test runs with m5a.large (2 cpus) x 20 nodes. - Full version runs with m5a.4xlarge (16 cpus) head node + 99 x m5a.large worker nodes - The smoke test takes **longer** than the full run due to syncing overhead at the end caused by the smaller head node instance (since all the syncing is going there). This PR bumps the smoke test's head node instance size, and forces trials to run on worker nodes -- the head node is purely meant to handle syncing in this test. This also fixes a problem that existed before, where the full 100 trial release test would schedule 8 trials on the head node, rather than utilize every node in the cluster. See ray-project#36346 for more context. Signed-off-by: Justin Yu <justinvyu@anyscale.com> 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?
There are 2 versions of the test: a smoke test version with 20 nodes/trials and a full version with 100 nodes/trials. The smoke test version timed out recently.
The test setups is slightly different:
This PR bumps the smoke test's head node instance size, and forces trials to run on worker nodes -- the head node is purely meant to handle syncing in this test. This also fixes a problem that existed before, where the full 100 trial release test would schedule 8 trials on the head node, rather than utilize every node in the cluster.
See #36346 for more context.
Questions
Some tests I ran (10 [session.report](http://session.report) calls, spaced out by 3 seocnds each → 30 seconds ideal run time)
m5a.large
head node (2 cpus):m5a.4xlarge
head node (16 cpus):Is this expected / ok performance for the super tiny head node? This most likely does not reflect real usage.
Related issue number
Closes #36346
Checks
git commit -s
) in this PR.scripts/format.sh
to lint the changes in this PR.method in Tune, I've added it in
doc/source/tune/api/
under thecorresponding
.rst
file.