Remote agent does not restart after server outage

Still need help?

The Atlassian Community is here for you.

Ask the community

This behavior has been fixed on Bamboo 6.10.2.

Symptoms

The remote agent's wrapper log (in <bamboo-agent-home>) shows:

FATAL  | wrapper  | 2010/01/29 21:54:09 | There were 5 failed launches in a row, each lasting less than 3 seconds.  Giving up.
FATAL  | wrapper  | 2010/01/29 21:54:09 | There may be a configuration problem: please check the logs.
STATUS | wrapper  | 2010/01/29 21:54:09 | <-- Wrapper Stopped

After that the remote agent has to be restarted manually.

Cause

The default configuration of the wrapper repeats an agent start-up attempt 5 times, each start-up attempt times out after 3s (that is if no successful connection could be made).

Resolution

For longer periods of server outages please configure the wrapper in <bamboo-agent-home>/conf/wrapper.conf with

# How long should the wrapper wait before it considers an invocation successful?  3 seconds should be long
# enough for any configuration errors to have been determined.
wrapper.successful_invocation_time=3 << keep this
wrapper.max_failed_invocations=5 << change this

Increasing the 'max_failed_invocations' parameter will increase the number of attempts.  

The total time before the agent fails is equal to (in seconds):

wrapper.restart.delay (default 5) * wrapper.max_failed_invocations


This would be the equivalent value 2h reconnect attempts after the server outage:

wrapper.successful_invocation_time=3
wrapper.max_failed_invocations=1440

Related improvement request:  BAM-18608 - Getting issue details... STATUS


Last modified on Dec 5, 2019

Was this helpful?

Yes
No
Provide feedback about this article
Powered by Confluence and Scroll Viewport.