Commit 9fab7eb
authored
feat(dashboards): add operational panels (#16229)
## Motivation
Add observability panels to Grafana dashboards covering operational
blind spots: xDS snapshot sizing, KDS health, VIP exhaustion, cert
expiry, injection outcomes, and DNS internals.
## Implementation information
**Control plane dashboard** (`kuma-control-plane.json`):
- xDS Snapshot Resources (p95 by type) — track cluster/endpoint fan-out
growth
- KDS Active Zone Connections — monitor zone CP connectivity
- KDS NACK Rate — surface config rejection issues
- VIP Allocation Exhaustion (1h) — alert before VIP pool runs out
- xDS Cert Time Remaining (per mesh) — track cert renewal health
- Sidecar Injection Outcomes — split success/skip/error injection counts
**Service debug dashboard** (`kuma-service-debug.json`):
- DNS Queries by Type & Source — break down DNS traffic patterns
- DNS Response Codes — surface NXDOMAIN/SERVFAIL spikes
- DNS Map Entries (per pod) — track DNS table size per dataplane
All panels use existing metrics exposed by kuma-cp and kuma-dp; no code
changes required.
> Changelog: feat(dashboards): add operational Grafana panels for xDS,
KDS, DNS, and cert monitoring
---------
Signed-off-by: Marcin Skalski <skalskimarcin33@gmail.com>1 parent 6d32503 commit 9fab7eb
2 files changed
Lines changed: 739 additions & 51 deletions
0 commit comments