Skip to content

Conversation

@toddpalino
Copy link

Described in #3436

This adds a configuration parameter timeout_delay which allows a user to specify a grace period for workers to start notifying that they are live. This decouples the time it takes a worker to start up from the time to detect a problem after it is running. The default does not change the existing behavior - if the configuration is not specified, no delay is added.

The delay works by adding time to the initial creation of the WorkerTmp notification file. The arbiter will evaluate the notification as good until the delay period expires. If the worker starts notifying before that period expires, it effectively cancels the rest of the delay.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant