Metrics reference

All services expose Prometheus metrics on a dedicated port configured via metrics.listen.

Edge Gateway metrics

Request metrics

Metric	Type	Labels	Description
eg_requests_total	counter	host, status, source	Total requests processed
eg_request_duration_seconds	histogram	host, source	Request processing time

Cache metrics

Metric	Type	Labels	Description
eg_cache_hits_total	counter	host, dimension	Cache hit count
eg_cache_misses_total	counter	host, dimension	Cache miss count
eg_cache_stale_total	counter	host, dimension	Stale cache served count

Render metrics

Metric	Type	Labels	Description
eg_render_requests_total	counter	host, status	Render requests sent
eg_render_duration_seconds	histogram	host	Render request duration
eg_render_errors_total	counter	host, error_type	Render errors

Bypass metrics

Metric	Type	Labels	Description
eg_bypass_requests_total	counter	host, status	Bypass requests
eg_bypass_cache_hits_total	counter	host	Bypass cache hits

Render Service metrics

Chrome pool metrics

Metric	Type	Labels	Description
rs_chrome_pool_size	gauge		Configured pool size
rs_chrome_pool_active	gauge		Currently active instances
rs_chrome_pool_available	gauge		Available instances

Render metrics

Metric	Type	Labels	Description
rs_render_total	counter	status	Total renders
rs_render_duration_seconds	histogram		Render duration
rs_render_errors_total	counter	error_type	Render errors

Instance lifecycle

Metric	Type	Labels	Description
rs_chrome_restarts_total	counter	reason	Chrome instance restarts
rs_chrome_requests_per_instance	histogram		Requests before restart

Cache Daemon metrics

Queue metrics

Metric	Type	Labels	Description
cd_queue_size	gauge	host, priority	Queue size by priority
cd_queue_due_now	gauge	host, priority	Entries ready to process

Processing metrics

Metric	Type	Labels	Description
cd_recache_total	counter	host, status	Recache operations
cd_recache_duration_seconds	histogram	host	Recache duration

Grafana dashboard queries

Request rate

rate(eg_requests_total[5m])

Cache hit ratio

sum(rate(eg_cache_hits_total[5m])) / sum(rate(eg_cache_hits_total[5m]) + rate(eg_cache_misses_total[5m]))

Chrome pool utilization

rs_chrome_pool_active / rs_chrome_pool_size

P99 render latency

histogram_quantile(0.99, rate(rs_render_duration_seconds_bucket[5m]))

Alert examples

High error rate

alert: HighRenderErrorRate
expr: rate(eg_render_errors_total[5m]) / rate(eg_render_requests_total[5m]) > 0.05
for: 5m

Chrome pool exhaustion

alert: ChromePoolExhausted
expr: rs_chrome_pool_available == 0
for: 1m

Low cache hit rate

alert: LowCacheHitRate
expr: sum(rate(eg_cache_hits_total[5m])) / sum(rate(eg_cache_hits_total[5m]) + rate(eg_cache_misses_total[5m])) < 0.5
for: 10m