SOLVED: C11 service shuts down right after startup

[prikala]

C11 service (11.0.11, virtual machine, windows server 2008R2) shuts down during service startup

One test server is acting weird.
The cognos service is restarted every night to keep things clean.

Sometimes the service does not start, it seems the server shuts down by itself just after it has started. In the p2pd_messages log can be seen something like this
============
[14.9.2018 3:21:34:690 EEST] 00000131 SystemOut                                                    O 2018-09-14T03:21:34.690+0300   INFO   STATU.com.cogno.pogo.servi.DispatcherServices   [Timer-42]            NA               The dispatcher is ready to process requests.
[14.9.2018 3:21:38:305 EEST] 000001a2 SystemOut                                                    O Configure Request Detected.
[14.9.2018 3:21:38:600 EEST] 000001ab com.ibm.ws.webcontainer.servlet                              I SRVE0242I: [search] [/search] [com.ibm.bi.search.SearchService]: Initialization successful.
[14.9.2018 3:21:44:860 EEST] 0000005e com.ibm.ws.app.manager.AppMessageHelper                      A CWWKZ0001I: Application glass-framework started in 340,982 seconds.
[14.9.2018 3:21:45:820 EEST] 0000005e com.ibm.ws.webcontainer.osgi.mbeans.PluginGenerator          I SRVE9103I: A configuration file for a web server plugin was automatically generated for this server at D:\cognos\c11_server\wlp\usr\servers\cognosserver\logs\state\plugin-cfg.xml.
[14.9.2018 3:21:54:775 EEST] 00000016 com.ibm.ws.kernel.launch.internal.ServerCommandListener      A CWWKE0055I: Server shutdown requested on 14. syyskuuta 2018 at 3:21. The server cognosserver is shutting down.


=========

As a workaround I have scheduled a repeating service start task after the restart task. The service will eventually start and stay up.

I guess this could be some kind of timing problem.
I wonder what is requesting server shutdown?

EDIT 2018-11-12: It seems the "net start" command is somehow messing with the cognos service. I replaced it with "sc start" since then Cognos seems to start reliably every time.

[sdf]

Is this just 1 server setup?
And your job/s "Restart" not "Stop" and "Start"?

[prikala]

This is single server setup.

The restart script is actually stop and start. Does that matter?
----
  net stop "IBM Cognos:9305"
  timeout 10
  net start "IBM Cognos:9305"
  timeout 20
----
I added timeouts in hope that things are just happening "too quickly".  No help.

The separate startup script has only the "net start" command (and "timeout" after that).

I guess this must be something that is caused by windows scheduler or "net".
Starting manually through Cognos Configuration has never failed.

[sdf]

That's what I was about to ask.
To increase the timeout after net stop.
something like :
timeout /t 120

you can also include a state check if the service is Running or Stopped
before executing the next command.

This are just my suggestions you can try.