Jobs Metrics
Metrician can automatically report metrics about your delayed_job, resque, or sidekiq job system.
Configuration
Metrician requires 2 configuration options be set in your copy of metrician.yaml to report jobs metrics. Timing and error metrics are enabled by default. For each job or error, the name of the job class in question can also be reported as part of the metric. The following are the default settings.
:jobs:
:enabled: true
:run:
:enabled: true
:error:
:enabled: true
:job_specific:
:enabled: false
Metrics
app.jobs.run |
the average time (ms) it takes for jobs to execute |
|---|---|
app.jobs.error |
the number of times jobs raise exceptions during execution |
app.jobs.run.job.<job_name> |
the average time (ms) it takes for specific types of jobs to execute |
app.jobs.error.job.<job_name> |
the number of times specific types of jobs raise exceptions during execution |
Jobs Metrics Query Language Expressions
The current job execution rate can be computed with the following expression on a graph:
gauge_count(app.jobs.run)
The total wall time jobs spend executing can be computed with the following expression:
gauge_total(app.jobs.run)
Job error rate can be computed with the following expression:
app.jobs.error / gauge_count(app.jobs.run)