Page MenuHomePhabricator

Handle it better when a managed gRPC server crashes on startup
ClosedPublic

Authored by dgibson on Tue, Nov 17, 1:14 AM.

Details

Summary

This should not be very common, but a user did hit it when the dagster.grpc module wasn't available. Instead of crashing and then waiting 15 seconds for a message that never comes, then timing out, detect right away that the process is no longer alive while waiting for the message.

Test Plan

New Test

Diff Detail

Repository
R1 dagster
Lint
Automatic diff as part of commit; lint not applicable.
Unit
Automatic diff as part of commit; unit tests not applicable.

Event Timeline

Harbormaster returned this revision to the author for changes because remote builds failed.Tue, Nov 17, 1:33 AM
Harbormaster failed remote builds in B21245: Diff 25772!

fearow

python_modules/dagster/dagster/serdes/ipc.py
131–143

nit: i confused myself because i forgot what exactly poll does - maybe rename this _check_for_process_exit or something

This revision is now accepted and ready to land.Tue, Nov 17, 3:53 PM
This revision was landed with ongoing or failed builds.Tue, Nov 17, 4:03 PM
This revision was automatically updated to reflect the committed changes.