Skip to content
View in the app

A better way to browse. Learn more.

Unraid

A full-screen app on your home screen with push notifications, badges and more.

To install this app on iOS and iPadOS
  1. Tap the Share icon in Safari
  2. Scroll the menu and tap Add to Home Screen.
  3. Tap Add in the top-right corner.
To install this app on Android
  1. Tap the 3-dot menu (⋮) in the top-right corner of the browser.
  2. Tap Add to Home screen or Install app.
  3. Confirm by tapping Install.

[Solved] High cpu load on only 2 cores? other at 0%

Featured Replies

Attached you will see my specs and an example of my cpu load any time I have anything running on plex. It seems like a phenom 2x4 840 would be powerful enough for this application. Im not sure if unraid keeps any configuration between computers, but this particular installation/usb was previously on a computer with similar specs, but a high end fx processor. Since then I've moved it to my media server and have been working to get all the functionality from that server working with unraid. Anyway, thats the history, if it matters.

 

The problem here is that the plex docker (linuxserver's repo) and indeed the entire system seem to be only using the first 2 cores of the processor. I have no vm's and unraid is not running as a vm itself, so there should be no reason for it to not use the entire cpu... unless this is a limitation of unraid, which would make me sad.

 

Does anybody have any advice on where to look?

specs.png

load.png

Edited by realmonster

That's plenty of proc to run plex.

 

I would load and run a docker netdata.  With this docker you can parse out the resources each docker is consuming.  I have two dockers running currently, \ The docker the hostnames are unconventional.  Mine are 009b8df5731b and dad2090c2f4c

 

Running the Web UI for netdata you can see exactly the resources each VM is consuming.  I would validate what your seeing on the unraid dashboard CPU metrics with what unraid is stating on the dashboard page.  

 

Since 6.3.2 I have an issue where the dashboard CPU stats simply quit working after a period of time, the percentages just hang at a current state that's not a true representation of actual load.  Using netdata to validate actual use. 

Here's the breakdown of CPU core utilization and the individual plex docker CPU resource requirement upon a fresh start of the docker using netdata

 

totalproc.pngdockerproc.thumb.png.0ba3f0b7adeac46995dd563b6d967f15.png

Edited by joedotmac

  • Author

Thank you for showing me that docker, very usefull troubleshooting tool. Unfortunately, if I'm reading this right, all I'm confirming is that it is indeed only using 2 cores. All the mac addresses I selected and screenshotted are mostly at 0 with a few little spikes, I don't know why I have 2 of them using cpu, I only have plex working on anything.

 

Im pretty sure a00beffc616 is plex.

 

Note that screenshot 1 is overall CPU utilization shortly after playing a 1080p movie with no subtitles, original quality and original sound playing through a different computer on firefox, sorry I chopped the nav bar off on that screenshot.

stress test.png

mac 1.png

mac 2.png

Edited by realmonster

Does the docker ever release the CPU or does it maintain that utilization level for hours?  Mine sometimes when I start it consumes a considerable amount of CPU for two to three minutes.  I believe it's Plex attempting to update library meta data. 

 

If your unsure which is Plex, stop Plex and watch which docker instance stops producing data into the graph.  I had to do the same.  

 

It would appear one docker is using 45% and the other, likely Plex is using 145% of a total 400% compute available.  

 

None of your other dockers exhibit this behavior?  I'm using binhex-plex FWIW, it's updated quite frequent. Not certain if anything points to the docker itself, but worth a mention.  

 

 

 

 

  • Author

Ill try out the binhex repo later tonight. Ive confirmed that a00beffc616 is plex and attached is that docker idleing. Also attached is my list of installed dockers, mostly low demand python stuff, but the main problem is my suspicion that only 2 cores are being used by plex in the first place. I understand that plex is the vast majority of processing requirement, but I was running pretty much this same config on steamos (debian fork if you didnt know) and had much better performance.

dockers.thumb.png.855009420a1f8c8d4cd3a222718cc657.pngidle.thumb.png.ea53002fe4c2af5c17b1027cb398909c.png

  • Author

Solved. In my original post I mentioned that this unraid usb was used in another computer with an fx processor. Somehow while in that environment isolcpu was set to 2,3,4,5,6,7 which locked those cores out of the unraid os. I simply used the GUI and went to Main>Flash and modified that line to not include 2,3 and now all 4 cores are being used by plex. Unfortunately its still using them 100%, but thats not unraids fault, im going to try overclocking to fix that problem.

 

Thanks for the help joedotmac as well as these two threads

https://forums.lime-technology.com/topic/54287-isolcpus-results-in-only-1-core-being-used/

https://www.tapatalk.com/topic/18593-unraid/50299-only-getting-30-performance-with-games-cant-determine-bottleneck

Archived

This topic is now archived and is closed to further replies.

Account

Navigation

Search

Search

Configure browser push notifications

Chrome (Android)
  1. Tap the lock icon next to the address bar.
  2. Tap Permissions → Notifications.
  3. Adjust your preference.
Chrome (Desktop)
  1. Click the padlock icon in the address bar.
  2. Select Site settings.
  3. Find Notifications and adjust your preference.