Skip to content

[DO NOT MERGE] ci: restore main validation and retry full nodes - #172

Open
limou102 wants to merge 5 commits into
mainfrom
dev/limou/fix_main_ci_ruff_260916
Open

limou102 wants to merge 5 commits into
mainfrom
dev/limou/fix_main_ci_ruff_260916

Conversation

@limou102

@limou102 limou102 commented Sep 16, 2026

Copy link
Copy Markdown
Collaborator

Description

The latest full CI run on main first fails in lint because Ruff reformats three projection files added by recent merges. After that formatting fix, the first PR run reaches GPU testing and e2e-mixed (sglang) lands on crsuse2-m2m-054, whose containerd storage is full.

This PR applies the exact Ruff output and treats no space left on device during a dispatched tier as a node failure: the harness excludes the named node and retries elsewhere within the existing five-submission limit.

Type of change

  • Bug fix (non-breaking change which fixes an issue)
  • Infra/Build change

Changes

  • Restore repository-pinned Ruff formatting in three projection files.
  • Retry a dispatched GPU tier away from a node with exhausted local storage.
  • Add a regression test for node exclusion and successful retry.

Checklist:

  • The functionality is complete
  • My changes generate no new warnings
  • I have added a test that proves the retry fix is effective
  • New and existing relevant unit tests pass locally with my changes

Validation

  • SKIP=gofmt pre-commit run --all-files --show-diff-on-failure
  • pytest -q tests/unit/projection (290 passed, 80 skipped)
  • pytest -q tests/unit/e2e_harness/test_slurm_retry_limit.py (12 passed)

Main reproduction: https://github.com/AMD-AGI/Infera/actions/runs/35077892039
First PR run: https://github.com/AMD-AGI/Infera/actions/runs/35078093915

Signed-off-by: LI MOU <lxglbk@gmail.com>
Copilot AI lite review requested due to automatic review settings September 16, 2026 09:12

Copilot AI left a comment

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

🟢 Approval recommended

Formatting-only changes were reviewed with no unresolved issues.

Pull request overview

Restores repository-pinned Ruff formatting across three projection files without changing behavior.

Changes:

  • Reformats projection test assertions and comprehensions.
  • Reformats validation error messages.
  • Resolves the reported CI lint failure.
File summaries
File Description
tests/unit/projection/test_tuning_disaggregated_pools.py Ruff formatting updates
tests/unit/projection/test_prefill_curve_extrapolation.py Ruff formatting update
infera/projection/agents/tuning_agent/inference_tuning.py Ruff formatting updates
Review details
  • Files reviewed: 3/3 changed files
  • Comments generated: 0
  • Review effort level: Lite

💡 Add a code-review agent skill or configure MCP servers for context-aware, tailored reviews. Learn more in the docs.

Signed-off-by: LI MOU <lxglbk@gmail.com>
@limou102 limou102 changed the title [DO NOT MERGE] style: restore projection ruff formatting [DO NOT MERGE] ci: restore main validation and retry full nodes Sep 16, 2026
Signed-off-by: LI MOU <lxglbk@gmail.com>
Signed-off-by: LI MOU <lxglbk@gmail.com>
Signed-off-by: LI MOU <lxglbk@gmail.com>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants