Query, compare and break down time series across services and environments.
1 hour window · 1m resolution
Summary per series over the selected window
95th percentile request latency.
Available metrics and their semantics
Aggregate CPU utilization across all running replicas.
Resident set size as a percentage of the container memory limit.
Inbound request rate measured at the service boundary.
Share of requests returning 5xx or failing at the application layer.
Median request latency.
99th percentile request latency. Tail behaviour under load.
Active connections to the primary database pool.
Number of jobs waiting to be processed.
Share of replicas passing health checks.
7 services · 10 metrics · 7 potential series