Where is the analysis report saved?

The skill creates a local '.work/prow-job-analyze-resource/' directory organized by Build ID, where it stores the downloaded logs and the final interactive HTML report.

Can I search for multiple resources at once?

Yes, the skill supports comma-separated lists and complex regex patterns (e.g., 'pod1|pod2') to match multiple resources across different namespaces.

What types of logs does this skill analyze?

It specifically parses Kubernetes API audit logs in JSONL format and plain-text pod logs found within the 'gather-extra' artifacts of a Prow job.

Does this skill require gcloud authentication?

No, the target Prow results bucket is typically public, so read access is available via gcloud without requiring manual authentication.

What happens if I analyze the same Build ID twice?

The skill detects existing artifacts and asks whether you want to use the local cache or re-download the logs, saving time and bandwidth on repetitive analyses.

Prow CI Resource Analyzer

Name: Prow CI Resource Analyzer
Author: openshift-eng

byopenshift-eng

•

デプロイメントとDevOps

Analyzes Kubernetes resource lifecycles in Prow CI jobs by parsing audit and pod logs to generate interactive timeline reports.

概要

This skill streamlines the debugging of Prow CI test failures by automatically extracting and analyzing Kubernetes resource events from Google Cloud Storage (GCS) artifacts. It parses complex audit logs and pod logs to reconstruct a chronological history of resource states, including creation, modification, and deletion. By providing interactive HTML reports with detailed timelines, it helps developers quickly identify where and why specific resources, such as pods, deployments, or configmaps, failed during ephemeral test cluster execution.

主な機能

Chronological event sorting with HTTP status code and glog level analysis
Generation of interactive HTML reports with expandable event details and timelines
Simultaneous parsing of JSONL audit logs and plain-text pod logs
Regex-based resource matching across specific namespaces and kinds
Automated artifact retrieval from GCS using gcloud CLI integration
29 GitHub stars

ユースケース

Auditing service account actions and API interactions within ephemeral test clusters
Debugging intermittent test failures in OpenShift or Kubernetes CI pipelines
Tracking the lifecycle of specific pods or deployments during an e2e test run

概要

主な機能

Chronological event sorting with HTTP status code and glog level analysis
Generation of interactive HTML reports with expandable event details and timelines
Simultaneous parsing of JSONL audit logs and plain-text pod logs
Regex-based resource matching across specific namespaces and kinds
Automated artifact retrieval from GCS using gcloud CLI integration
29 GitHub stars

ユースケース

Auditing service account actions and API interactions within ephemeral test clusters
Debugging intermittent test failures in OpenShift or Kubernetes CI pipelines
Tracking the lifecycle of specific pods or deployments during an e2e test run