Genesys CTI User Forum
Genesys CTI User Forum => Genesys CTI Technical Discussion => Topic started by: PFCCWA on January 05, 2015, 10:05:38 AM
-
Hello,
We are using statserver 8.1.2 and have over a short period experienced similar outages of the application without finding a cause.
The symptoms are typically that it freezes yet sci/ga is showing as started and running.
In addition the app has stopped writing logs (unfortunately it was at standard level before issue).
Resolution is to restart it.
It has a major impact on platform and operations because there is a failover to backup (see point about sci/ga showing application as started/running).
Clients such as URS/data sourcer/ccpulse stop working because statistic requests produce errors - here is typical one (opening error: server is closed or doesn't exist). Also there is no alarm or alert to indicate issue.
At a guess it could have run out of resource due to large statistic requests but has been running without issue for a while.
Host OS is unix.
Has anyone experienced similar problem?
Can the core dump file give us any answers? im not sure if we can analyse this or genesys support will have tools.
thanks.
-
The core dump is very helpful in similiar issues. According to my experience with SS component, the latest "working" version of StatServer is 8.1.1xx.xx. We have encountered many issues in higher version and Genesys support itself has recommended downgrade in some cases.
-
Just as general points...
Remove any unused stats from the SS Application in CIM/CME. Take a look at what you need for your Services and remove everything else.
Consider using more instances - one per Service - and it is defo not a good idea to be using one instance of SS with Routing [i]and[/i] Reporting...
Check that your local instance of LCA is unaffected - it might not be responding which would lead to you thinking SS is running, when it isn't.
Set up an Alarm/Reaction to autostart SS, after a graceful shutdown.
HTH?
AG
-
hello,
I should have been more clear, we have dedicated statservers for routing and reporting (rtime and hist).
Its a good point about LCA, I believe this crashed during investigation and attempts to restart the SS app so might answer question about why we did not see its failure.
Only other thing is increase log level to debug although it may not reoccur..
thanks,
-
It makes sense that, if LCA is down, SCS/SCI won't report that the SS has gone down. Extrapolating your comments - are you sure the box is stable? I mean, if LCA and SS are [i]both [/i]going down...
I've had issues in the past with regular backups and security/anti-virus sweeps stopping Genesys Services. If it happens regularly, check "when" it occurs and look to see if anything else on the box is set to run/start concurrently. If you can't stop the regular maintenance schedules, look at gracefully stopping and starting the Services, around those scheduled times. If you find your scheduled Services are the cause - try to move them to a more convenient time (overnight, if you're not 24/7)...
HTH?
AG
-
Hey,
we had a similar problem a while back, and here is what we found (the hard way):
- statserver really stops working well if you get more than 200K of stats being tracked in it
- it freezes when writing into statserver backup file (ssbackup.000) (go ahead, test it!)
Since a very long time ago, we made it a practice to disable or at least limit ssbackup: http://www.sggu.com/smf/index.php?topic=1516.0