Set traefik resource requests so HPAs can scale #520

Merged
benvin merged 1 commits from benvin/traefik-resource-requests into main 2026-10-04 15:27:50 +11:00
Member

Both traefik HPAs report ScalingActive False: missing request for cpu because the chart values set no container resources, so the CPU-utilisation metric can't be computed and neither Deployment ever scales.

  • traefik-external requests 200m CPU / 128Mi, memory limit 512Mi
  • traefik-internal requests 500m CPU / 256Mi, memory limit 1Gi
  • No CPU limits, to avoid throttling the proxy
  • The chart derives GOMEMLIMIT (90% of the memory limit) from the new limits
Both traefik HPAs report `ScalingActive False: missing request for cpu` because the chart values set no container resources, so the CPU-utilisation metric can't be computed and neither Deployment ever scales. - traefik-external requests 200m CPU / 128Mi, memory limit 512Mi - traefik-internal requests 500m CPU / 256Mi, memory limit 1Gi - No CPU limits, to avoid throttling the proxy - The chart derives GOMEMLIMIT (90% of the memory limit) from the new limits
unkin-agent added 1 commit 2026-10-04 15:21:33 +11:00
Set traefik resource requests so HPAs can scale
ci/woodpecker/pr/pre-commit Pipeline was successful
ci/woodpecker/pr/kubeconform Pipeline was successful
9b7af0221a
Author
Member
  • apps/overlays/au-syd1/traefik-system/values-internal.yaml:76 — cpu request 500m vs observed ~480m/pod (kubectl top) = ~97% utilisation against the 80% HPA target, so it scales up on first sync (2→3+) → size the request from real usage so steady state sits well below 80% (e.g. 1 cpu), or state in the PR body that the immediate scale-up is intended.
  • nit: values-external.yaml:76 — request values are unexplained in the body (observed ~105m cpu / 65Mi per pod) → add the sizing basis to the body.
- apps/overlays/au-syd1/traefik-system/values-internal.yaml:76 — cpu request 500m vs observed ~480m/pod (kubectl top) = ~97% utilisation against the 80% HPA target, so it scales up on first sync (2→3+) → size the request from real usage so steady state sits well below 80% (e.g. 1 cpu), or state in the PR body that the immediate scale-up is intended. - nit: values-external.yaml:76 — request values are unexplained in the body (observed ~105m cpu / 65Mi per pod) → add the sizing basis to the body.
benvin merged commit 9375052d6c into main 2026-10-04 15:27:50 +11:00
benvin deleted branch benvin/traefik-resource-requests 2026-10-04 15:27:50 +11:00
Sign in to join this conversation.
No Reviewers
No Label
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: unkin/argocd-apps#520