Skip to content

GitLab

"mmdet3d/apis/__init__.py" did not exist on "5e5b822bb0b4ec02639d3e8e750944a6024dc29f"
Unverified Commit 6806c4e6 authored by Xiaoyu Zhang's avatar Xiaoyu Zhang Committed by GitHub
Browse files

[CI monitor] Improve CI analyzer: fix job failure tracking and add CUDA-focused filtering (#11505)

parent 0c0779d6
Expand all Show whitespace changes
Inline Side-by-side
Showing with 76 additions and 554 deletions
scripts/ci_monitor/README.md
View file @ 6806c4e6
This diff is collapsed.
scripts/ci_monitor/ci_analyzer.py
View file @ 6806c4e6
......@@ -67,15 +67,14 @@ class SGLangCIAnalyzer:
return all_runs[:limit]
def analyze_ci_failures(self, runs: List[Dict]) -> Dict:
"""Analyze CI failure patterns"""
print("Analyzing CI failure data...")
"""Analyze CI failure patterns (CUDA jobs only)"""
print("Analyzing CI failure data (CUDA only)...")
# SGLang specific job categories
# SGLang specific job categories (CUDA only)
job_categories = {
"sgl-kernel": [
"build": [
"build-test",
"sgl-kernel-build-wheels",
"sgl-kernel-unit-test",
"sgl-kernel-mla-test",
],
"unit-test": [
"unit-test-frontend",
......@@ -87,11 +86,35 @@ class SGLangCIAnalyzer:
"performance": [
"performance-test-1-gpu-part-1",
"performance-test-1-gpu-part-2",
"performance-test-1-gpu-part-3",
"performance-test-2-gpu",
],
"accuracy": ["accuracy-test-1-gpu", "accuracy-test-2-gpu"],
"deepep": ["unit-test-deepep-4-gpu", "unit-test-deepep-8-gpu"],
"b200": ["unit-test-backend-4-gpu-b200"],
"accuracy": [
"accuracy-test-1-gpu",
"accuracy-test-2-gpu",
],
"mla-test": [
"sgl-kernel-mla-test",
],
"deepep": [
"unit-test-deepep-4-gpu",
"unit-test-deepep-8-gpu",
],
"per-commit": [
"per-commit-8-gpu-h20",
],
"nightly": [
"nightly-test-perf-text-models",
"nightly-test-eval-text-models",
],
"integration": [
"run-all-notebooks",
"vllm-dependency-test",
"test-disaggregation",
],
"b200": [
"unit-test-backend-4-gpu-b200",
],
}
stats = {
......@@ -141,13 +164,26 @@ class SGLangCIAnalyzer:
job_name = job.get("name", "Unknown")
job_conclusion = job.get("conclusion", "unknown")
# Filter out non-specific CI jobs
if job_name not in [
# Filter out non-specific CI jobs and non-CUDA jobs
# Skip meta jobs and AMD/NPU related jobs
if (
job_name
not in [
"check-changes",
"pr-test-finish",
"pr-test-h20-finish",
"pr-test-amd-finish",
"pr-test-b200-finish",
"lint",
]:
"Set up job",
]
and "-amd" not in job_name.lower()
and "mi300" not in job_name.lower()
and "mi325" not in job_name.lower()
and "gfx" not in job_name.lower()
and "-npu" not in job_name.lower()
and "ascend" not in job_name.lower()
):
# Record successful jobs (update last success)
if job_conclusion == "success":
stats["job_last_success"][job_name] = {
......@@ -158,7 +194,7 @@ class SGLangCIAnalyzer:
}
# Record failed jobs
elif job_conclusion == "failure" and run_status == "failure":
elif job_conclusion == "failure":
stats["job_failures"][job_name] += 1
# Store failure link (keep only last 3 for each job)
......@@ -216,7 +252,7 @@ class SGLangCIAnalyzer:
return pr_info
def _analyze_failure_pattern(self, job: Dict, stats: Dict):
"""Analyze failure patterns"""
"""Analyze failure patterns (CUDA jobs only)"""
job_name = job.get("name", "")
steps = job.get("steps", [])
......@@ -224,19 +260,33 @@ class SGLangCIAnalyzer:
if step.get("conclusion") == "failure":
step_name = step.get("name", "")
# SGLang specific failure pattern recognition
# SGLang specific failure pattern recognition (CUDA only)
if "timeout" in step_name.lower():
stats["failure_patterns"]["Timeout"] += 1
elif "test" in step_name.lower() and "unit" in job_name.lower():
elif "build" in step_name.lower() or "build" in job_name.lower():
stats["failure_patterns"]["Build Failure"] += 1
elif "install" in step_name.lower() or "dependency" in job_name.lower():
stats["failure_patterns"]["Dependency Installation Failure"] += 1
elif "unit" in job_name.lower() or "unit-test" in job_name.lower():
stats["failure_patterns"]["Unit Test Failure"] += 1
elif "performance" in job_name.lower():
elif "performance" in job_name.lower() or "perf" in job_name.lower():
stats["failure_patterns"]["Performance Test Failure"] += 1
elif "accuracy" in job_name.lower():
stats["failure_patterns"]["Accuracy Test Failure"] += 1
elif "build" in step_name.lower():
stats["failure_patterns"]["Build Failure"] += 1
elif "install" in step_name.lower():
stats["failure_patterns"]["Dependency Installation Failure"] += 1
elif "mla" in job_name.lower():
stats["failure_patterns"]["MLA Test Failure"] += 1
elif "deepep" in job_name.lower():
stats["failure_patterns"]["DeepEP Test Failure"] += 1
elif "nightly" in job_name.lower():
stats["failure_patterns"]["Nightly Test Failure"] += 1
elif "notebook" in job_name.lower():
stats["failure_patterns"]["Notebook Test Failure"] += 1
elif "disaggregation" in job_name.lower():
stats["failure_patterns"]["Disaggregation Test Failure"] += 1
elif "h20" in job_name.lower() or "h200" in job_name.lower():
stats["failure_patterns"]["H20/H200 GPU Failure"] += 1
elif "b200" in job_name.lower():
stats["failure_patterns"]["B200 GPU Failure"] += 1
elif "gpu" in job_name.lower():
stats["failure_patterns"]["GPU Related Failure"] += 1
else:
......@@ -245,7 +295,7 @@ class SGLangCIAnalyzer:
def generate_report(self, stats: Dict):
"""Generate CI analysis report"""
print("\n" + "=" * 60)
print("SGLang CI Analysis Report")
print("SGLang CI Analysis Report (CUDA Only)")
print("=" * 60)
# Overall statistics
......
Markdown is supported
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment