Master how Cloud Run Jobs run batch workloads to completion.
0 / 5 completed
1 / 5
At standup, a dev needs a container that runs to completion, not one that serves HTTP. Which Cloud Run primitive fits?
A Cloud Run job runs a container to completion for batch or scheduled work, unlike a service which serves long-lived HTTP requests. Jobs are ideal for ETL, migrations, or one-off processing. This is the right choice when there is no request to respond to.
2 / 5
During a design review, the team wants a job to process 100 shards in parallel. Which setting controls this?
A Cloud Run job splits work into a number of tasks, and parallelism controls how many tasks run simultaneously. Each task reads its index via an environment variable to pick its shard. Tuning parallelism balances speed against resource quotas.
3 / 5
In a PR review, a dev references the single unit of work inside a job. What is it called?
A task is one instance of the job's container, identified by a task index. A job's execution runs the configured number of tasks. Designing idempotent tasks lets failed tasks retry safely.
4 / 5
During a code review, someone wants the job to run nightly. Which GCP service triggers it on a schedule?
Cloud Scheduler is GCP's managed cron service and can trigger a Cloud Run job execution on a schedule. You point it at the job's run endpoint with the right credentials. This pairs naturally with jobs for periodic batch work.
5 / 5
An incident report shows a single run of a job that launched several tasks. What is that run called?
An execution is one run of a Cloud Run job, which in turn launches the configured tasks. Each execution has its own status and logs. Distinguishing the execution from individual tasks is key when debugging partial failures.