[🤖] feat(k8s-system-coredns): add cache hit rate and forward request latency panels

[raman1236]

Summary

Add two new panels to the CoreDNS dashboard that surface derived metrics currently not visible, placed directly below the existing cache section.

New Panels

1. CoreDNS - Cache Hit Rate %

Shows the percentage of DNS queries served from cache:

sum(rate(coredns_cache_hits_total{...}[$__rate_interval]))
/
(
  sum(rate(coredns_cache_hits_total{...}[$__rate_interval]))
  +
  sum(rate(coredns_cache_misses_total{...}[$__rate_interval]))
) * 100

Why: The existing "Cache Hits / Misses" panel shows raw hit/miss rates, but operators need the derived ratio to evaluate cache effectiveness and tune cache-size and TTL settings. A drop in cache hit rate is a key signal for DNS performance degradation.

2. CoreDNS - Forward Request Latency

Shows p99 and p50 upstream DNS resolution latency:

histogram_quantile(0.99, sum by (le) (
  rate(coredns_proxy_request_duration_seconds_bucket{proxy_name="forward", ...}[$__rate_interval])
))

Why: The existing dashboard shows forward request count but not latency. When CoreDNS forwards queries to upstream resolvers, latency directly impacts DNS resolution time for pods. This panel helps diagnose slow upstream resolvers, network issues, or resolver overload.

Changes

• Added 2 new timeseries panels at y=43, below the cache hits/misses and cache size panels
• Shifted heatmap panels (DNS request duration, request size, response size) down by 8
• Dashboard version bumped from 22 to 23

Checklist

• Dashboard version bumped
• Conventional commit format used
• AI-assisted contribution clearly identified with 🤖 scope

[raman1236]

Friendly ping — this PR adds cache hit rate and forward request latency panels to the CoreDNS dashboard. Would appreciate a review when convenient. Thanks!