Disclaimer: I don't believe the server restart actually caused the issues, just possibly uncovered issues that were laying dormant.
I recently load balanced our two presence servers and rebooted them because frankly, it was time. With almost 900 users on it and no one could remember the last time we had. Prior to the reboot, all users were on one node, and the secondary node was there as a failover. I split the difference, rebooted the servers. Logged in on my own and made sure it was fine and called it a night. The next morning, the nightmare begun. Here is what seems to be happening:
Previous versions of the program (10.6.x) will not sign in at all. Updating to 12.0.1 helps usually, but sometimes, that won't stick. Sometimes, after every PC restart or sign out with J4W, the user can no longer login. The client has to be reset and then signed back in again.
When we installed 10.5 we had copied over a jabber-config.xml file locally. Now it is stored on the tftp server and I don't know if some remnants of the local file are fighting with the server file being pulled down. I just a moment ago uninstalled the new version on a users PC, deleted all AppData folders pertaining to Jabber and restarted their PC. Reinstalled Jabber 12.0.1. Worked perfectly. Signed them out. Signed back in. Same issue as before (under options, accounts, cannot connect to IM server anymore). Only a reset works, and that only lasts as long as they stay signed in.
I'm checking through the config file now to see if we messed something up somewhere, but didn't know if this was a bug or if we just goofed something up.
Bonus round: One user, just one, cannot be seen as available no matter what. Our Directory URI is set as our mail address, but her profile will not seem to pull away from FirstInitialLastName@domain even though we have checked and double checked CUCM and AD. Her contacts hub shows her email address up top instead of her name. Servers communicate fine between themselves. I have purged every inkling of Jabber from her PC and reinstalled and this still happens. Any ideas?
I am close to my knowledge limit here on what is the cause. Checking configs and contemplating moving all users back to a single node like we had before is what I have planned at the moment.
