Seiri guides

Background job monitoring

Background job monitoring is about workers that should keep ticking. When the last successful check-in is too old, Seiri alerts you — whether the worker crashed, stalled, or the scheduler forgot the job.

Stacks that fit

Background workers fail quietly when the process dies, the queue stalls, or a recurring task stops being enqueued. Seiri watches for check-ins — not CPU graphs.

• ·Sidekiq / Resque / Delayed Job — ping after each successful job or on a worker heartbeat interval. Guide: monitor Sidekiq.
• ·Celery / RQ — Beat and workers check in when periodic tasks finish. Guide: monitor Celery.
• ·Bull / BullMQ — emit a heartbeat when a repeatable job completes successfully.
• ·Hangfire / Quartz / Spring @Scheduled — same push model: success path pings Seiri.
• ·Custom workers in Go, Node, or Python — one HTTP call or email heartbeat at the end of the unit of work.

Pattern

Emit a heartbeat at the end of each successful unit of work (or on a fixed interval for long-lived workers). Pair with grace periods so brief jitter doesn’t page you. This is heartbeat monitoring — not website uptime and not a cron-webhook scheduler.

# After the worker finishes useful work
curl -fsS -m 10 --retry 3 https://ping.seiri.app/webhook/acme-demo:YOUR_REF/success

For Kubernetes Deployments and Jobs, prefer seiri-kube-agent HealthCheck CRs so silence in-cluster still alerts.