[Bug] Endless cycle during heavy load

[rix1337]

Current Behavior

The startup scripts are overzealous with this specific check:

[supervisor  ] starting service 'xvnc'...
[supervisor  ] service 'xvnc' failed to be started: not ready after 5000 msec, giving up.
[supervisor  ] stopping service 'xvnc'...

My system is low power and very able to run JDownloader.
It happens frequently that during container startup this error message appears, killing and restarting the container increasing the system load even more.

Expected Behavior

Increase timeout or don't kill container at all.

Steps To Reproduce

Have high system load / simulate high system load.
Try to boot the container normally

Environment

• OS: Unsaid
• OS version: 7.0.0
• CPU: Intel J4105
• Docker version:
• Device model:
• Browser/OS:

Container creation

Default settings from UNRAID Community apps

Container log

[supervisor  ] starting service 'xvnc'...
[supervisor  ] service 'xvnc' failed to be started: not ready after 5000 msec, giving up.
[supervisor  ] stopping service 'xvnc'...

Container inspect

Anything else?

No response

[jlesage]

Just to clarify: will the container eventually start correctly ?

I will add the ability to configure the timeout via a variable.

[rix1337]

Yep - it always will
Under load the startup might just go through 10 cycles - depending on when the load eventually drops.

I think exposing the timeout as a parameter is all I need
Thanks!

[jlesage]

Default timeout has been increased from 5 to 10s. If it's not enough, you can use the DEFAULT_SERVICE_READY_TIMEOUT environment variable to increase it (value in msec).

[rix1337]

Awesome thx