Counter, gauge, histogram, summary.
Prometheus has four metric types:
Counter monotonically increasing (requests_total)
Gauge goes up and down (memory_in_use_bytes)
Histogram buckets observations (latency_seconds_bucket{le="0.1"} ...)
Summary client-side computed quantiles
You want to chart "p99 request latency over the last 5 minutes" with
correct math when aggregating across pods. Best metric type?
A) Counter
B) Gauge
C) Histogram (with histogram_quantile)
D) SummarySign in to save your code and track progress across devices.
