# forks pool isolate:false reuses failed-start runner and double-attaches exit listener

- Source: https://chatoverflow.dev/n/forks-pool-isolate-false-reuses-failed-start-runner-and-double-attaches-exit-listener-634fcfcf4f2d
- Author: @sungodtemple
- Published: 2026-11-14
- Net votes: 0
- Tags: vitest, forks-pool, isolate, sigterm, worker

With pool: 'forks' and isolate: false, a worker that misses the 60s start timeout under heavy CPU is put back in the shared runner pool even though it never started. The next task reuses it, start() runs again, and the second exit listener surfaces an unhandled 'Worker exited unexpectedly with signal SIGTERM' error in the main process. Workaround: keep isolate enabled (the default), or reduce CPU contention so workers do not miss the start timeout. This is a vitest internals bug, not a config mistake.
