SolarWinds.JobEngineWorker.x64.v2 consuming all CPU

Suddenly, I was blind... or at least if feels that way because Orion/NPM/SAM won't serve anything but a log in page because one of four SolarWinds.JobEngineWorker.x64.v2 processes is consuming all available CPU. it's been doing that for about an hour now. It's done this before and it eventually stops, but it would be nice to know what's going on. Any ideas?

Usually, people add pictures, so here's mine.

Find more posts tagged with

sam 6.3.0

npm 12.0.1

Accepted answers

All comments

mloomis

Somehow it occurred to me to be less specific and search for CPU issues. I came across this article which may be applicable.

NPM - CPU Maxed Constantly - False Latency Alerts

I changed

Small Rx Buffers from "Not Present" to 64 and

Rx Ring #1 Size from "Not Present" to 32.

VMware showed me that these events happen about every 12 hours. We shall see.

mloomis

I caught it in the act. I incrementally changed

Small Rx Buffers from 64 to 2048 and

Rx Ring #1 Size from 32 to 2048. In the middle of the last change, the CPU hogging process stopped(?) and CPU went back down to a more normal 60%. So, was the change in CPU utilization due to changing the receive buffers... OR due to the process timing out after 1 hour? My hunch is on the the latter, but I have no idea what caused that process to run for one hour at that time.

pstranak

Hi,

is the command line of the problematic process always the same or different?

mloomis

I don't know. I'm a network guy, but I dug around and found what you're talking about. Next time I catch it misbehaving I'll report what I find. Thanks!

mloomis

I just removed DPA integration & NTA (both of which I didn't have support on and were never set up properly to begin with) and upgraded to NPM 12.1 and SAM 6.4, so I don't know if that will help fix this issue or create new ones, or both.