Practise the standard verbs for configuring Airflow DAG retry policies.
0 / 5 completed
1 / 5
Fill in: 'We ___ a bounded retry count with exponential backoff on every task so a transient network blip doesn't retry forever, spamming the same failure.'
We 'set a retry count' — the standard, simple collocation for configuring how many times a task attempts before failing. The other options are less idiomatic here.
2 / 5
Fill in: 'Leaving retries unlimited on a task calling a broken API can ___ the scheduler stuck retrying the same failing task all night.'
We say unlimited retries will 'leave' the scheduler stuck retrying — the standard, natural collocation for the resulting problem. The other options aren't idiomatic here.
3 / 5
Fill in: 'We ___ a task-level timeout alongside retries so a hung external call doesn't block the whole DAG run from completing.'
We 'configure' a timeout — the standard collocation for setting up a bound on how long a task may run. The other options aren't idiomatic here.
4 / 5
Fill in: 'We ___ on the final retry failure only, not every individual attempt, so on-call isn't paged for a transient blip that self-resolved.'
We 'alert' on a failure — the standard, simple collocation for triggering a notification from a specific condition. The other options are less idiomatic here.
5 / 5
Fill in: 'We ___ the retry history in the Airflow UI after a DAG run to tell a genuine outage apart from an intermittent flaky dependency.'
We 'inspect' history — the standard, simple collocation for reviewing a record of past attempts. The other options are less idiomatic here.