Skip to main content

One post tagged with "observability"

View All Tags

Continuous Profiling on AKS with Pyroscope, Blob Storage, and Managed Grafana

· 20 min read
Diego Casati
Diego Casati
Principal Cloud Architect, Azure Global Black Belt
Post Updates

2026-05-20 — Updated based on lessons learned from a live deployment:

  • Removed hardcoded pyroscope.image.tag from values-azure.yaml to prevent chart/image version mismatches when the chart is upgraded
  • Added pyroscope.extraLabels with azure.workload.identity/use: "true" to propagate the label to all pod templates (the chart uses extraLabels, not podLabels)
  • Pinned --version 2.0.1 in the helm upgrade --install command
  • Added a Troubleshooting callout documenting the two most common crash patterns and their fixes

You deploy your workloads on AKS and collect metrics with Prometheus and logs with Loki. But when latency spikes hit, you stare at dashboards knowing something is slow without knowing where in your code the time is being spent.

That's the gap continuous profiling fills.