I attempted an upgrade to 2025.4.3 last week and experienced multiple issues. The first was that during the main poller's configuration wizard, I received the following error.
Error while executing script- The index 'idx_voicdrs_orgphon_id' is dependent on column 'DateTimeDisconnectUTC'.
The index 'idx_voipcdrs_destphoneid_id_pkid' is dependent on column 'DateTimeDisconnectUTC'.
ALTER TABLE ALTER COLUMN DateTimeDisconnectUTC failed because one or more objects access this column.
I created a ticket with support and in the meantime found an existing KB with a similar error that referenced dropping the index and rerunning the wizard. As we were getting ready to do this support replied and also stated that was the fix. Our DBA ran it against our DB and we were able to move on. The configuration wizard continued and we then hit a snag when it was updating settings for Application Templates.
I found an article for this (https://solarwindscore.my.site.com/SuccessCenter/s/article/Configuration-stuck-in?language=en_US) and ended up getting it to continue on using resolution 2 (killing the process) since changing the max tasks didn't help. Once the wizard was complete, support suggested we run it again to confirm it could complete, which it did.
Our main poller and website was up, so I went ahead and used centralized upgrade to update the APEs. All was successful and I rebooted the main poller and all APEs so they were clean. All services started up as expected and there were no errors in the event log.
When I jumped back into Solarwinds to confirm all modules were working I found that WPM was completely broken. The two transaction player servers we had were showing 'location not found' and were showing down under Node Management, even though they were up and pinging. I had found that for some reason, they were both hanging on agent upgrade. I tried multiple things to get them to succeed - rebooting both servers, manually uninstalling agent then manually installing, but nothing worked. I went to look at the agent status and found that our "Manage Agents" page would just spin and then eventually time out. I rebooted the main poller again and gave it some time and tried again, but again it timed out.
The end of the change window was coming up so I made the decision to roll back, so I grabbed a diagnostic from the main poller and went back to 2024.4.1. The rollback was successful. Unfortunately support says without the logs on the WPM player they don't really have much to go by so the plan is to attempt 2025.4.3 in two weeks and grab diagnostics from every poller just in case its needed.
I thought it was odd that the agent wouldn't update as we have went from various versions of Solarwinds (2018, 2020, 2022, 2024) without any problems with agents updating or WPM having trouble communicating. I know that there is a port change in 2026.1, but found no documentation about any changes in 2025.4.x. The only agents that didn't update were on the player servers. They were on the same subnet as other agents that updated just fine.
QUESTIONS:
Has anyone ran into anything I experienced with WPM?
What would cause the automatic agent updater to hang?
Even if the agent wasn't installed, shouldn't the player been able to connect to the WPM service?
Does 2025.4.3 require a newer .net version? We are currently on 8.0.12 on all Solarwinds servers.
