-
Notifications
You must be signed in to change notification settings - Fork 1.6k
ci: Fix L0_batch related flaky tests #7999
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
base: main
Are you sure you want to change the base?
Conversation
# and the remaining requests will form the second batch. | ||
# and the remaining requests will either form the second batch | ||
# or more batches depending on their arrival time. |
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.
Can you explain why the originally intended behavior is not consistent and needs to be relaxed here?
@@ -1895,7 +1897,7 @@ def test_preferred_batch_only_use_no_preferred_size(self): | |||
def test_max_queue_delay_only_non_default(self): | |||
# Send 12 requests with batch size 1. The max_queue_delay is set | |||
# to non-zero. Depending upon the timing of the requests arrival | |||
# there can be either 1 or 2 model executions. | |||
# there can be either 1 or multiple model executions. |
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.
Can you explain why the originally intended behavior is not consistent and needs to be relaxed here? (different for zero and non-zero queue delay scenarios?)
@@ -619,13 +619,15 @@ def test_multi_batch_not_preferred_different_shape(self): | |||
}, | |||
) | |||
) | |||
# Add some delay to ensure the first two requests arrive before the third | |||
time.sleep(2) |
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.
Should this 2
seconds be some function of the _max_queue_delay_ms
instead?
… yinggeh-DLIS-7826-fix-L0-batcher
What does the PR do?
Set looser time requirement and execution counts to L0_batcher* tests. The actual response time and execution counts are highly dependent to the request arrival time.
Checklist
<commit_type>: <Title>
Commit Type:
Check the conventional commit type
box here and add the label to the github PR.
Related PRs:
Where should the reviewer start?
Test plan:
L0_batcher--base
L0_batcher_shm--base
L0_batcher_cudashm--base
23748413
Caveats:
Background
Related Issues: (use one of the action keywords Closes / Fixes / Resolves / Relates to)