[Test] Fix test_debug_dump_no_args timeout in remote-server mode#9230
Merged
Conversation
The test consistently times out in --remote-server mode on Buildkite CI. The test framework auto-prepends `sky status -u` before the actual test command. In deploy mode (--deploy), the API server's on-boot `sky check` blocks the SHORT worker pool while validating cloud credentials (~38s). This blocking is intentional — other commands depend on credential discovery. With a 30s timeout the test always fails before sky check completes. Increase to 120s to match the other debug-dump test. Fixes: https://buildkite.com/skypilot-1/smoke-tests/builds/9448 Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Collaborator
Author
|
/smoke-test --kubernetes --remote-server -k test_debug_dump_no_args |
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
2 participants
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Summary
test_debug_dump_no_argstimeout from 30s to 120s (matchingtest_debug_dump_nonexistent_resources)Root Cause Analysis
The test fails 100% of the time in
--remote-server --kubernetesnightly builds (7/7 retries): https://buildkite.com/skypilot-1/smoke-tests/builds/9448Investigation (SSH debug on Buildkite agent):
sky status -ubefore the actual test command--remote-servermode, the nested Docker container starts the API server withsky api start --deployburstable_parallelism=0(by design, to prevent OOM in production)sky.checkas a SHORT request to discover cloud credentials (~38s for all clouds)sky.check+ 3 daemon requests (status-refresh, volume-refresh, managed-job-refresh) saturate the SHORT worker poolsky status -ugets queued but cannot be scheduled untilsky.checkfinishesWhy the blocking is intentional:
sky.checkon boot validates cloud credentials. Until it completes, the server doesn't know which clouds are available. Other commands depend on this credential discovery, so blocking the SHORT pool during boot is correct behavior — it prevents serving requests with incomplete cloud state.Why this wasn't a problem before:
test_debug_dump_no_argswas newly introduced in #8868. The 30s timeout was set without accounting for the deploy-mode boot sequence in--remote-serverCI. The on-boot blocking has always existed.Test plan
test_debug_dump_nonexistent_resources)sky status -ucompletes in ~38s in deploy mode, well within 120s🤖 Generated with Claude Code