SWJobEngineWorker2 crash

Question

The SolarWinds platform's SWJobEngineWorker2.exe process is experiencing unexpected shutdowns, resulting in large memory dump files (.dmp) (some up to 1,200 MB). These files have caused disk C: to fill up on the affected servers.

The problem occurs recurrently on two specific servers: the Main Poller and an Additional Polling Engine.

Any idea what could be happening?

chad.every · Answer

There is an issue fixed in 2026.2.2 (coming soon), that fixes an issue with the job worker when nodes, polled via WMI, have WMI related issues (access/auth, TCP port, etc). It causes an issue that would attempt to spin up more job engine worker processes, creating a memory leak.

Unsure if this is your exact scenario, but I'd look at any WMI nodes on the two servers. Potentially move WMI nodes to a single engine (if possible) to see if the problem becomes isolated to that engine. If it does, then you'd want to look into the logs to find out the offending WMI nodes and/or wait to upgrade)

mjonathan · Answer

we are in the SolarWinds Observability Self-Hosted Advanced Enterprise Scale version: 2025.1.1

We have not tried running the configuration wizard, we open a ticket with support and they tell us that by analyzing logs the process ends due to an uncontrolled memory exception (System.OutOfMemoryException), linked to the KERNELBASE.dll module

They suspect possible security software interference (McAfee and CrowdStrike). They requested to apply specific SolarWinds directory exclusions and also pointed out the possibility of database pressures (SQL deadlocks, lack of cleaning temporary files)

cnorborg · Answer

Have you tried running the config wizard?  Tends to get rid of most issues…   You should also probably let us know what version you're using and such…