tinkr metrics

Query a metric time series from a service.

tinkr metrics <service> [options]

Arguments

Argument	Description
service	Service name as configured in the active profile

Options

Flag	Default	Description
--metric TEXT	http_requests_total	Metric name to query
--since TEXT	1h	Look-back window — e.g. 30m, 2h, 24h
--json	off	Emit raw JSON

Examples

# HTTP request rate (last hour)
tinkr metrics payments-api

# P99 latency
tinkr metrics payments-api --metric latency_p99 --since 2h

# Error rate over 24h
tinkr metrics payments-api --metric error_rate --since 24h

# Raw JSON for scripting
tinkr metrics payments-api --metric cpu_usage --since 30m --json

Output

Metric: http_requests_total  Service: payments-api  Window: 1h

  14:00  ████████████████████ 1240 req/min
  14:05  ████████████████████ 1318 req/min
  14:10  ████████████████████ 1201 req/min
  14:15  ██████████░░░░░░░░░░  620 req/min   ← drop
  14:20  ████░░░░░░░░░░░░░░░░  230 req/min

Common metrics by backend

Grafana / Prometheus

http_requests_total
http_request_duration_seconds
process_cpu_seconds_total
go_memstats_heap_alloc_bytes

CloudWatch

RequestCount
Latency
5XXError
CPUUtilization

Datadog

trace.web.request
trace.web.request.duration
aws.ecs.cpuutilization

Azure Monitor

requests/count
requests/duration
requests/failed

See also

• tinkr anomaly — automatically detect unusual metric behavior
• tinkr diff — compare metric values between two time windows