schedulix 2.11 server not running stable

[domini...@conrad.de]

Dear schedulix masters,

we have some strange behaviours on schedulix 2.11 server.

After recreating our server and starting the server, we are in some start -> error -> fail - > retry loop.

-rw-r--r--  1 schedulix schedulix  21596 Sep  5 10:04 BICserver.out.178
-rw-r--r--  1 schedulix schedulix  14019 Sep  5 10:06 BICserver.out.179
-rw-r--r--  1 schedulix schedulix  18733 Sep  5 10:09 BICserver.out.180
-rw-r--r--  1 schedulix schedulix  13491 Sep  5 10:11 BICserver.out.181
-rw-r--r--  1 schedulix schedulix  21688 Sep  5 10:13 BICserver.out.182

I also see this error in the logfile

[scrolllog] Waiting for child (384903) to terminate
[scrolllog] Child exited with state 143
[scrolllog] Try to restart child (child terminated with exit code <> 0)

We have a service in the background, which will start the server and here we see this restart counter.
schedulix-server.service: Scheduled restart job, restart counter is at 12.

We start our server with this command:
$BICSUITEHOME/bin/server-start

Also for me it looks too less information, even when I switched TraceLevel=3

Can you please have a look, what is going on?

Best regards

Dominik

[Ronald Jeninga]

Hi Dominik,

if you use a service in the background to restart the scheduling server, you're basically duplicating scrolllog functionality, which will restart the server if it exits with an exit code unequal zero.

This again can lead to a number of servers running at the same time, which then will quarrel about who's the owner of the repository.

So either you should tell your service not to restart or tell your service to monitor the scrolllog process instead (that could crash as well, albeit unlikely).

The log file you've added just shows normal operation. Both the start messages and the failure messages are missing.

Maybe you can post one of the others?

Anyway, before doing so, please try to find out how many servers are running.

If you've multiple instances, kill all of them, together with the scrolllog processes.

Your service should then do a server-start again and you should be back to normal.

If my suspicion is correct, there's no need to post the other log file any longer, since we've then found the cause.

Best regards,

Ronald

[domini...@conrad.de]

Hi Ronald,

after stopping the service and starting the schedulix server manually, it looks good.

Thanks for your hint.

"Anyway, before doing so, please try to find out how many servers are running."

We only saw 2 processes running for BICserver.
/home/schedulix/schedulix/bin/scrolllog ...
and 
/usr/bin/java ... de.independit.scheduler.BICServer

And it was the same on our Live environment.
Test environment has schedulix 2.11

Live environment has schedulix 2.9

With this information we thought, that there is a problem with schedulix itself or we did something wrong with our migration on Test env.

We haven't thought that there could be two or more running instances of schedulix.

Best regards,

Dominik

[Ronald Jeninga]

Hi Dominik,

great! I'm happy your issue is resolved now.

Thank you for the feedback!

Enjoy your weekend

Ronald