Skip to content

Guides

How to handle long-running jobs

Some jobs run for hours and occasionally hit the six-hour limit.

5 min read

Some jobs run for hours and occasionally hit the six-hour limit.

Why it happens

Work that should be sharded is running serially, and the default job timeout is too permissive to catch a hang.

How to fix it

  1. Shard the work across a matrix so each leg finishes in a reasonable time
  2. Set timeout-minutes deliberately on every job
  3. Checkpoint long work to an artifact so a retry does not start over
  4. Move genuinely long batch work out of CI and onto scheduled infrastructure

Your next build could be twice as fast, at half the price

Start free. Migrating away is the same one line, and we publish that diff too.