Planning an upgrade, has anyone recently upgraded from SolarWinds Observability Self-Hosted 2024.4.1 to 2025.4.1. If so, any known issues you encountered?
@u303083 I know folks who have done the upgrades without any issues. But personally, I haven't done so far.
I've got a couple.
RabbitMQ - I have seen this with a couple of versions moving to a 2025.x. Allegedly this was corrected, but be aware. RabbitMQ local settings database on polling engines can become corrupted with an in-place upgrade.
Symptoms: Lots of them. Failed config wizard, HA pool will not start, many others. In the RabbitMQ log file you will see a lot of errors, and the main one I have seen is a login failed for user 'Orion'
You will need to stop RabbitMQ service then rename or delete the data folder located in {OSDrv/Usually C:}:\ProgramData\Solarwinds\Orion\RabbitMQ. It's best to run a wizard after this
HA- As mentioned above, the RabbitMQ issue can cause your HA pool to malfunction. Be aware, also, that if you have any Orion Remote Collectors in your instance, the update can cause a bug that will prevent the "My Deployment" page from working. The fix is to remove the ORC and re-add it. From our support case, this one has been passed to development for investigation.
UDP ephemeral ports- Sometime between 2023.4.2 and 2025.2.0, Job Engine v2 began using the UDP ephemeral range for SNMP communication. Our security software picked this up in my primary environment. I have access to another environment where I found that this had also happened, so be aware that if you have a security whitelisting procedure, you may run into this (although it could have already happened in 2024.4.1). This is NOT documented in the Solarwinds port requirements yet, but it is supposed to be added soon.
We did an in place upgrade to 2025.4.0 on 11/17/25 without any issues.
From 2024.4.1?
I upgraded last week. I ran into some issues pre-upgrade, and a few issues post upgrade. However, in all fairness, I believe my issues were with our environment and not with SolarWinds itself.
Pre-upgrade - We have 4 additional web servers, HA main polling servers and 14 additional polling engines. The upgrade guild said to stop the SolarWinds services on all systems before the upgrade. That in of itself causes nothing but problems, in our environment. I ended up rebooting all of the servers and leaving all SW services running, but I left HA disabled. Once I did that the upgrade without any hiccups.
Post upgrade, some of my custom SWQL tables and queries broke. I had complaints of issues with both classic views and modern dashboards, but most of the issues self-healed within 24 hrs…., which was the oddest thing I've ever seen. I have a few odd custom SWQL classic widgets that aren't happy, but those have been easy to fix.
My biggest issue is, I've had to uninstall/reinstall WPM player & recorder software on all 10 WPM servers. Plus, I had to rebuild the agent data settings and the SEUM Agent Domain Configuration. Again, not a big deal and is most likely something 100% isolated to our environment.
good to know about the services! I also read that too to stop services prior.
I had a question as this is my first upgrade. I have HA as well. When I kick off the update tool from the web console does it detect all of my servers? Active/standby core, active/standby pollers and web servers?
we face lot of issue when we update our non-prod from 2024.4.1 to 2025.4.0.
We also did the upgrade and the only issue I remember is needing to run the config wizard on one polling engine.
To be 100% honest I don't worry that much about if all the systems being detected pre-install on the update tool. I figure I can always download the installers and install them after the main polling engine is upgraded.
I mainly focus on the post installation and making sure all of the polling engines are up and operating correctly.
