-
Type:
Suggestion
-
Resolution: Handled by Support
-
None
-
Component/s: None
-
None
Hello,
I am maintaining a Bamboo Remote Agent in a company where Bamboo server is maintained centrally by administrators. The Agent is run on Windows as a service. I have no admin access to Bamboo server itself, I can only request for service.
I often myself in a situation where the Remote Agent is unavailable, has been for some time, and I need to manually restart it. There are special capability requirements for the agent, and these are not provided by the other Agents.
According to our Bamboo administrators, there is a timeout for situations when network connection is lost. After the timeout, the Remote Agent stops retrying and shuts down. I see this as a problem. Remote Agent is a backend software, there is not necessarily anyone there to handle the restart. This can cause longer disruption to service than is really necessary.
One can work around this problem by configuring the (windows) service to be restarted automatically. But that will then happen in all error scenarios: configuration errors, Bamboo bugs, etc. And in these cases I would argue that the correct response is to shut down like it currently does. I.e. two error categories would be required. If Remote Agent can recover from error, it should try to do it infinitely. If it cannot recover, it should shut down.
Now, it may be the case that I am misinformed, and this is the way Remote Agent already functions. Maybe there is configuration option or something that I am unaware of. In that case, just let me know, and I'll pass the information along to our admins.
Thanks for your help and feedback,
-Marko