how to gracefully restart QlikView server with batch file?

[daveatkins]

I need to schedule a restart of qlikview and wrote a batch script to do this, but when I scheduled it to run nightly, it failed to restart the server.

The script is scheduled to run as the service account user with highest permissions and from windows task scheduler, reports success. Does anyone have any suggestions? Or a more complete script you have written for the same purpose?

The delays are built in because sometimes it can take the qlikview server 4-5 minutes to shut down if it has grown to be using 100-200GB of RAM and it is important for all services to have ended before attempting restart as that will lead to locking and need for a hard reboot. Maybe I just need to trap the distribution service also? Here is my script:

rem script to restart qlikview every night

sc stop QlikviewDistributionService
sc stop QlikviewManagementService
sc stop QlikviewServer

rem wait 15 seconds and then check for successful stop of QlikviewServer
:recheck
ping 127.0.0.1 -n 15 -w 1000 > nul

sc query QlikviewServer | find /I "STATE" | find "STOPPED"
if errorlevel 1 goto :recheck
goto :start

:start
sc start QlikviewServer
ping 127.0.0.1 -n 5 -w 1000 > nul
sc start QlikviewDistributionService
ping 127.0.0.1 -n 5 -w 1000 > nul
sc start QlikviewManagementService

[Josh_Berg_Support]

The bigger question here is why you're restarting the Qlikview Server service every night, what is the reason for this?  In order for Qlikview to perform as expected, it needs cached result sets stored in QVS, restarting the service will clear this cache and will require it to be rebuilt.  

Are you experiencing an issue that requires QVS to be restarted nightly?  If so, can you please explain or provide log entries showing your issue?  What version of Qlikview are you running?  Is this an existing or recent issue after a change in your Qlikview environment such as an upgrade?

Thanks,
Josh

Qlik Support

[daveatkins · Accepted solution]

Josh, thank you for commenting...this is a continuation of the issue we had a ticket on. I will reply to that email thread. the restart has not solved the problem; we had another failure today just a few hours after restarting