Repository navigation

Add diff, multi-account, spot, K8s GPU metrics, Prometheus - #26

Merged

maksimov merged 39 commits into

masterfrom

feature/k8s-gpu-metrics

Apr 21, 2026

Merged

Add diff, multi-account, spot, K8s GPU metrics, Prometheus#26
maksimov merged 39 commits into
masterfrom
feature/k8s-gpu-metrics

Commits on Apr 19, 2026
- Add K8s GPU metrics collection design spec
  maksimov
  committed
  7a2802e
- Add K8s GPU metrics collection implementation plan
  maksimov
  committed
  d271797
- Add EnrichK8sGPUMetrics for CloudWatch Container Insights GPU metrics
  maksimov
  committed
  1c2d3d8
- Add ProxyGet to K8sClient interface for pod API proxy
  maksimov
  committed
  96155c1
- Add DCGM exporter scraping for K8s GPU metrics
  maksimov
  committed
  84a4a9b
- Add Prometheus query enrichment for K8s GPU metrics
  maksimov
  committed
  0f460c4
- Add ruleK8sLowGPUUtil for utilization-based K8s GPU waste detection
  maksimov
  committed
  d605cb4
- Wire K8s GPU metrics fallback chain into CLI scan flow
  maksimov
  committed
  54dc0ce
- Fix DCGM node matching and CW error spam
  maksimov
  committed
  c93e0f7
- Fix DCGM scrape spam and Prometheus node name mismatch
  maksimov
  committed
  2640b88
- Include time window in low GPU utilization recommendation text
  maksimov
  committed
  b08c025
- Skip CW enrichment when AWS creds unavailable, reduce DCGM noise
  maksimov
  committed
  e846cc8
Commits on Apr 21, 2026
- Add diff package with Compare function and tests
  maksimov
  committed
  ff5b7f5
- Add diff table and JSON output formatters
  maksimov
  committed
  1eed3b8
- Add diff subcommand to compare two scan results
  maksimov
  committed
  5672229
- Fix box alignment in diff table output
  maksimov
  committed
  2e95784
- Fix misleading idle duration in K8s GPU node recommendations
  maksimov
  committed
  01b4512
- Update README with K8s scanning, diff command, and current output format
  maksimov
  committed
  c9ee92d
- Add multi-target scanning design spec
  maksimov
  committed
  cd27862
- Add multi-target scanning implementation plan
  maksimov
  committed
  b8c6a35
- Add TargetSummary and TargetErrorInfo model types for multi-target scanning
  maksimov
  committed
  408399e
- Extract BuildSummary to summary.go and add BuildTargetSummaries
  maksimov
  committed
  e856adb
- Implement ResolveTargets with STS AssumeRole for multi-account scanning
  maksimov
  committed
  d1c79f3
- Refactor Scan() for parallel multi-target scanning
  maksimov
  committed
  cd49f91
- Add --targets, --role, --org, --external-id, --skip-self flags to scan command
  maksimov
  committed
  698b0f6
- Add per-target summary table and target column to table formatter
  maksimov
  committed
  19fbbb7
- Add per-target summaries to markdown and Slack formatters
  maksimov
  committed
  8737a44
- Add cross-account and Organizations permissions to iam-policy output
  maksimov
  committed
  68fbeaa
- Add multi-account scanning docs to README
  maksimov
  committed
  1f58d92
- Fix callerAccount bug, deduplicate severity logic, clean up dead code
  maksimov
  committed
  6001598
- Add SpotHourlyCost field to GPUInstance model
  maksimov
  committed
  41b0867
- Implement EnrichSpotPrices with DescribeSpotPriceHistory
  maksimov
  committed
  b081710
- Wire EnrichSpotPrices into scanRegion after EC2 discovery
  maksimov
  committed
  82ae997
- Correct spot instance cost using live spot prices
  maksimov
  committed
  8f4973e
- Add ruleSpotEligible analysis rule for spot recommendations
  maksimov
  committed
  6b7bfab
- Add ec2:DescribeSpotPriceHistory to IAM policy output
  maksimov
  committed
  4b8c1c6
- Address review: update signal type comment, add pagination note, guard div-by-zero
  maksimov
  committed
  1abda17
- Add Prometheus GPU metrics support for EC2 instances
  maksimov
  committed
  2f05b9c
- Sanitize example output in README
  maksimov
  committed
  0c33990

Commits

Commits on Apr 19, 2026

Commits on Apr 21, 2026