wagoe.jobs.shell.worker — com.wagoe/wagoe-jobs 1.1.0-rc-1

Background job worker implementation.

Workers poll job queues, execute jobs, and handle retries. This module provides a production-ready worker with:

• Graceful shutdown
• Configurable polling intervals
• Automatic retry handling
• Scheduled job processing
• Job handler registry
• Comprehensive error handling

create-job-registry clj

(create-job-registry)

Create a new job handler registry.

Returns: JobRegistry implementing IJobRegistry

create-worker clj

(create-worker config queue store registry)

Create and start a background job worker.

Args: config - Worker configuration map with: :queue-name - Queue to process (keyword, required) :poll-interval-ms - Milliseconds between polls (default 1000) :scheduled-interval-ms - Check scheduled jobs interval (default 5000) :virtual-threads? - Run the loop on a virtual thread (default false) :tenant-runner - (fn [tenant-id f]) that calls f in the tenant's schema; a job with :metadata :tenant-id fails without one :metrics, :error-reporter - where each run's outcome goes (optional, see wagoe.jobs.shell.outcomes) :on-outcome - (fn [outcome]) called after each run (optional) :report-failures-every-ms - error-tracker throttle (default 1h) queue - IJobQueue implementation store - IJobStore implementation registry - IJobRegistry implementation

Returns: Worker implementing IJobWorker

create-worker-pool clj

(create-worker-pool config queue store registry)

Create multiple workers for parallel job processing.

Args: config - Worker pool configuration with: :queue-name - Queue to process :worker-count - Number of workers (default 4) :poll-interval-ms - Polling interval :scheduled-interval-ms - Scheduled check interval queue - IJobQueue implementation store - IJobStore implementation registry - IJobRegistry implementation

Returns: Vector of Worker instances

failure clj

(failure e)

What a job failure stores and logs: no value a database driver quoted, since it lands in the jobs table (BOU-633). :type names the kind of failure, the thrown class has its own key (BOU-323).

fatal? clj

(fatal? t)

Whether t is an error no job should swallow: the JVM may not survive it. Out of memory, an internal VM error, a killed thread. A StackOverflowError has unwound by the time it is caught, so it is the handler's failure, as an AssertionError or ExceptionInInitializerError is (BOU-834).

run-pending! clj

(run-pending! config queue store registry)

Run every job that is due, on the calling thread, until no queue has one left. Returns how many ran. For tests: with a worker thread the outcome is a race (BOU-469). A retry or a scheduled job lands in the future, so this ends.

stop-worker-pool! clj

(stop-worker-pool! workers)

Stop all workers in a pool gracefully.

Args: workers - Vector of Worker instances

worker-pool-status clj

(worker-pool-status workers)

Get status of all workers in a pool.

Args: workers - Vector of Worker instances

Returns: Vector of worker status maps