08-23-2011 09:49 AM - edited 03-14-2019 08:25 AM
Hello, I am looking for some help trying to figure out what changes are made after moving to a HAoWAN cluster in relation to DB reporting for the wallboard app.
I had/have the free comminuty wallboard script (thank you comminuty) working just great before the upgrade to a HA cluster. I am using UCCX 8.0.2SU3 and I have just reenctly added a second node for redundancy. My stats for the wallboard app seem to have stopped working the moment I started the upgrade. I have done the steps outline in the "Using Wallboard Software in a High Availability (HA) Deployment" guide but I am still not seeing current data.
Has anybody else had this problem or have any suggestions on a workaround?
John P
Solved! Go to Solution.
08-23-2011 10:51 AM
Hi
Well... not sure which wallboard/version of that wallboard you are using, so no idea if it checks that API.
However.. If you think you have it pointed at the pub, and the pub is master, then I would check that the actual tables are updating:
Try:
run uccx sql db_cra select * from rtcsqssummary
That should hopefully get you a list of the CSQs and stats; if it's not current/updating regularly (run it a few times on each server and compare) then the stats may not be updating and you need to look at the UCCX server.
If you see them updating on the pub, then you know you have a problem with the wallboard.
Aaron
08-23-2011 01:36 PM
Hi John
No problem - I would probably start by going into the UCCX admin web pages, and disabling the realtime stats snapshot. Re-enable it, and see if it starts up.
I did once have a problem where this stalled, and a restart of a server fixed it up. Can't honestly recall whether I had to apply a 'proper' fix to resolve it long term.
Regards
Aaron
08-23-2011 09:52 AM
Hi John
As far as I'm aware nothing has changed with HAoWAN (compared to old fashioned HA) and nothing has changed between v7 and 8.
In older versions, the rtCSQSummary and the other rt table are ONLY updated on the current 'Master' server. Your wallboard needs to be able to read from both and determine which one has updated recently, or be able to detect the current live server in some other way.
A lot of wallboards require manual intervention after a failover - i.e. reconfigure the DSN to point at the other box.
Regards
Aaron
08-23-2011 10:12 AM
Thanks Aaron, I will double check the tables that are being checked and make sure I have the correct table selected.
Cheers,
08-23-2011 10:21 AM
Hi John
You will have the correct tables (this will be coded in the app) it's just whether you are pointed to the right server.
Regards
Aaron
08-23-2011 10:28 AM
08-23-2011 10:51 AM
Hi
Well... not sure which wallboard/version of that wallboard you are using, so no idea if it checks that API.
However.. If you think you have it pointed at the pub, and the pub is master, then I would check that the actual tables are updating:
Try:
run uccx sql db_cra select * from rtcsqssummary
That should hopefully get you a list of the CSQs and stats; if it's not current/updating regularly (run it a few times on each server and compare) then the stats may not be updating and you need to look at the UCCX server.
If you see them updating on the pub, then you know you have a problem with the wallboard.
Aaron
08-23-2011 01:33 PM
Aaron, after runing the sql statement a few times on each server I can see that the table is not getting updated. It is showing the "stale data" that I am seeing on the wallboard app/script webpage. Thank you for that at least I know can start looking else where for my problem.
Thank you for sticking with me on this issue.
Cheers,
John
08-23-2011 01:36 PM
Hi John
No problem - I would probably start by going into the UCCX admin web pages, and disabling the realtime stats snapshot. Re-enable it, and see if it starts up.
I did once have a problem where this stalled, and a restart of a server fixed it up. Can't honestly recall whether I had to apply a 'proper' fix to resolve it long term.
Regards
Aaron
08-24-2011 11:06 AM
Aaron, I removed the snap shot config from the the pub and saw that the sub also recieved the settings. Then I entered the setting back, and they also showed up on the sub. But after runing the sql select statement I was still getting stale data, I opend a TAC case and they did a test in the lab with the same results. Like you suggested the real work around is a cluster reload. I will have to wait a coulple of weeks for change controls but I am sure that will be the end of it.
Keep up the great support.
Cheers.
Discover and save your favorite ideas. Come back to expert answers, step-by-step guides, recent topics, and more.
New here? Get started with these tips. How to use Community New member guide