Background Jobs

Background jobs are a strong fit for Chirpier because the model is simple and operational. Each job run produces a small number of flat events that map directly to the signals you care about.

Recommended events

Event	Value	Why
`job.completed.count`	`1` per completion	Track throughput and spot processing stalls
`job.retry.count`	`1` per retry	Catch jobs that are failing and retrying repeatedly
`queue.backlog.gauge`	Current queue depth	Monitor queue pressure in real time
`job.duration_ms`	Duration in ms	Detect slow jobs before they block the queue

Logging events

JavaScript


await client.log({ agent: "worker.billing", event: "job.completed.count", value: 1,
  meta: { job_type: "invoice_generate", queue: "billing" } });
 
await client.log({ agent: "worker.billing", event: "job.duration_ms", value: 3200,
  meta: { job_type: "invoice_generate" } });
 
await client.log({ agent: "worker.billing", event: "queue.backlog.gauge", value: 42,
  meta: { queue: "billing" } });

Python


client.log(Log(agent="worker.billing", event="job.completed.count", value=1,
    meta={"job_type": "invoice_generate", "queue": "billing"}))
 
client.log(Log(agent="worker.billing", event="job.duration_ms", value=3200,
    meta={"job_type": "invoice_generate"}))
 
client.log(Log(agent="worker.billing", event="queue.backlog.gauge", value=42,
    meta={"queue": "billing"}))

Go


_ = client.Log(ctx, chirpier.Log{Agent: "worker.billing", Event: "job.completed.count", Value: 1,
    Meta: map[string]any{"job_type": "invoice_generate", "queue": "billing"}})
 
_ = client.Log(ctx, chirpier.Log{Agent: "worker.billing", Event: "job.duration_ms", Value: 3200,
    Meta: map[string]any{"job_type": "invoice_generate"}})
 
_ = client.Log(ctx, chirpier.Log{Agent: "worker.billing", Event: "queue.backlog.gauge", Value: 42,
    Meta: map[string]any{"queue": "billing"}})

Recommended first policies

Event	Period	Aggregate	Condition	Threshold	Why
`job.retry.count`	`hour`	`sum`	`gt`	`20`	Jobs failing and retrying too often
`queue.backlog.gauge`	`minute`	`max`	`gt`	`500`	Queue building up faster than workers can drain
`job.duration_ms`	`day`	`p95_est`	`gt`	`30000`	Slow jobs that might block the queue
`job.completed.count`	`hour`	`sum`	`lt`	`10`	Processing stall — too few completions

Using comparisons

Compare this hour’s retry count to the same hour yesterday to see if retries are normal or elevated:


const analytics = await client.getEventAnalytics(retryEventID, {
  view: "window", period: "1h", previous: "previous_1d",
});
// value_pct_change of +300% means retries are 3x higher than usual

Destination routing

Use destination policy_ids to separate job alerts by team. For example, route billing job alerts to a #billing-ops Slack channel and general infrastructure alerts to #platform-ops.