This discussion has been locked. The information referenced herein may be inaccurate due to age, software updates, or external references.
You can no longer post new replies to this discussion. If you have a similar question you can start a new discussion in this forum.

Enabled automatic updates .. lost dozens of agents.

So this isn't a new issue but I'm just getting around to addressing it . Upgraded to SEM 2020.4.1 a while back. We always disable automatic agent updates before an upgrade now due to a memory leak issue with a previous version of the SEM agent. I recently re-enabled automatic agent updates to get our 2020.2.1 agents up to 2020.4.1. As a result of this, our number of used workstation licenses went from about 212 to 147, so about 65 or so agents just completely lost connectivity with the appliance.

Any idea why this might have happened?

Any suggestions on easily / quickly remedying this situation?

  • I had a similar problem specifically with a small subset of Windows Server 2008R2 agents when going to 2021.2.1. Not all, but some. The log kept filling with this error: 

    Nio exception: An exception occurred during the execution of select(): java.io.IOException: An operation was attempted on something that is not a socket

    Support could find no definitive root cause. I've had to turn off automatic updates and downgrade those specific agents back to 2020.4.1. They are reporting fine. No newer agent versions are functioning on those nodes.

    I have a feature request open to have the ability to turn off updates on a node-by-node basis.

    thwack.solarwinds.com/.../disable-automatic-upgrades-on-individual-nodes