feat(macrobenchmarks): collect and report system metrics

[zhixiangli]

Fetch per-pod system metrics (CPU, memory, network) from Cloud Monitoring during macrobenchmarks. Reduce these metrics to the bottleneck pod (maximum peak and mean across pods) and include them in the final summary.

• Add metrics.monitoring to fetch metrics using Cloud Monitoring API.
• Update metrics.calculate to perform the reduction.
• Update schema and raw store to handle the new metrics.
• Update scrape_metrics.sh to run the collection.
• Add tests for the new functionality.

[gemini-code-assist]

Code Review

This pull request introduces system metrics collection (CPU, memory, and network) from Cloud Monitoring for macrobenchmarks. It adds a new monitoring module to fetch these metrics, updates the database schema and calculations to aggregate per-pod metrics to the bottleneck pod, and updates the scrape script to collect them on a best-effort basis. The review feedback suggests improving the robustness of RFC3339 datetime parsing in the monitoring module by handling lowercase timezone designators.

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 89.68%. Comparing base (4ce07fe) to head (4c3298d).
⚠️ Report is 2 commits behind head on main.

Additional details and impacted files

@@           Coverage Diff           @@
##             main     #952   +/-   ##
=======================================
  Coverage   89.68%   89.68%           
=======================================
  Files          16       16           
  Lines        3579     3579           
=======================================
  Hits         3210     3210           
  Misses        369      369           

☔ View full report in Codecov by Harness.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.