Skip to content
This repository has been archived by the owner on Jul 24, 2021. It is now read-only.

Tirex-backend-manager spins with 100% CPU if all backends die #4208

Open
openstreetmap-trac opened this issue Jul 23, 2021 · 1 comment
Open

Comments

@openstreetmap-trac
Copy link

Reporter: amm
[Submitted to the original trac issue database at 11.49pm, Thursday, 2nd February 2012]

If all backend processes die (with non restartable errors) and no more processes are running without tirex-backend-manager having received a sighup, then tirex-backend-manager sits there spinning with 100% CPU usage.

If all backends exited with non-restartable errors, tirex-backend-manager should rather terminate as well than use up 100% CPU.

@openstreetmap-trac
Copy link
Author

Author: stephankn
[Added to the original trac issue at 6.29am, Sunday, 13th July 2014]

How to reproduce?

The restart code could remember the timestamp of a process restart. If restarts happen too frequent, either sleep between or abort as something seriously broken which won't recover.

Sign up for free to subscribe to this conversation on GitHub. Already have an account? Sign in.
Projects
None yet
Development

No branches or pull requests

1 participant