fix(metrics): scope failed_updates to last day and add group label

[Harsh63870]

Summary

Fixes #1567.

nebraska_failed_updates previously counted every failed update event ever recorded and only exposed an application label. That made it hard to use for current rollout health.

This PR:

• Limits counts to failed update-complete events from the last day (same window as other reporting queries)
• Adds a group label (via instance_application → groups; missing group → unknown)
• Resets the gauge vector before each refresh to avoid stale series

Test plan

• cd backend && go test ./pkg/api/ -run TestGetFailedUpdatesMetrics
• Scrape /metrics and confirm series look like nebraska_failed_updates{application="...",group="..."}
• Confirm old failures outside the 1-day window are not counted

Notes

Adding the group label changes the metric’s label set. Existing Grafana queries that only use application can keep working with sum by (application) (nebraska_failed_updates).

[Copilot]

Copilot was unable to review this pull request because the user who requested the review has reached their quota limit.