prometheus_metrics

Prometheus Metrics Example

This example shows how to:

• Wire core.TaskSchedulerConfig.Metrics to the Prometheus exporter.
• Poll runner/pool Stats() into Prometheus gauges.
• Expose /metrics for scraping.

Run

go run ./examples/prometheus_metrics

While it is running, scrape metrics:

curl -s http://127.0.0.1:2112/metrics | grep '^taskrunner_'

The demo keeps the server alive for about 2 seconds.

What to look at

• taskrunner_task_duration_seconds: task execution latency histogram.
• taskrunner_task_panic_total: count of panics during task execution.
• taskrunner_task_rejected_total: count of rejected tasks.
• taskrunner_queue_depth: scheduler/runner queue depth.
• taskrunner_runner_pending: pending tasks from RunnerStats.
• taskrunner_runner_running: running tasks from RunnerStats.
• taskrunner_pool_queued: queued tasks from PoolStats.
• taskrunner_pool_active: active tasks from PoolStats.
• taskrunner_pool_delayed: delayed tasks from PoolStats.

PromQL examples

sum(rate(taskrunner_task_duration_seconds_count[5m])) by (runner)

max(taskrunner_queue_depth) by (runner)

taskrunner_pool_active{pool="metrics-pool"}