Hey all. I'm looking for some advice on setup. I'm reading everything I can to learn as much as possible since I'm new to all this.

 

This is the unit i use unRAID on: https://www.terra-master.com/global/f4-5147.html It's a Terra Master unit. It didn't lockup much when I ran TOS5/6 operating system. It glitched a lot and required OS reinstalls and other stuff broke, but wouldn't generally run high CPU like this. 

 

I ran the 2 1tb nvmes as raid 1 just in case one fails. 

 

I've attached everything I can think of. It seems when a few things happen, such as a couple of downloads, the CPU spikes and things run hot. Then everything basically slows down to a stop.

 

I believe I setup the cache pool correctly and I changed the docker appdata / qbittorrent downloads paths to /mnt/cache instead of user to use the cache pool.

 

You will see tdarr in one of the images. I do not use tdarr. Even just running 1 file basically took the system to a stop, so I uninstalled it.   

 

On the CPU pinning page I don't recall ticking those spots. I'm honestly not sure what it means, so that might be default. Unless I checked it and forgot. 

 

 

 

 

 

 

 

 

Edited December 30, 20241 yr by tenletters

your processor limited. to much running at once. even with docker pinning.

use another machine for plex applicaton the libray can stay on unraid...  

Which makes sense. It just seems odd this all ran on the TOS operating system without maxing CPU and that's a garbage OS. This os a kickass OS. It feels like I have something off. 

3 hours ago, bmartino1 said:

your processor limited. to much running at once. even with docker pinning.

use another machine for plex applicaton the libray can stay on unraid...  

https://www.terra-master.com/global/alltos/

terra OS may have kerneal level optimizations for that hardware. unraid has a striped down version of kerenal 6 for slackware. 

in widnows therms this is akin to installing the chipset drivers. Unraid kerenal may be pulling a generic intel version which is not 3rd party optimised.
But thats a different rabbit hole and without a plugin to install driveers and config settings...
 

Quote

It didn't lockup much when I ran TOS5/6 operating system. It glitched a lot and required OS reinstalls and other stuff broke, but wouldn't generally run high CPU like this. 

Based on TOS and your issues before its was also over taxed.

its not a bad cpu.
https://www.intel.com/content/www/us/en/products/sku/212322/intel-celeron-processor-n5095-4m-cache-up-to-2-90-ghz/specifications.html

 

i dont' have a good recomendation due to the number of concurent runnign docker...

 

Reserve CPUs for the Host:

Ideal with unraid: you would Leave at least 1-2 cores (preferably low-numbered ones) unpinned for Unraid's base OS and essential services.

 

Not sure of a bios or other settings that may or shold be enabled on that nas box:
 

Balance Across Cores:

You Spread workloads across physical cores to prevent overheating or throttling.

If you have hyperthreading enabled, ensure you're utilizing real cores over virtual ones for heavy tasks.
 

Isolate Critical Services:

Dedicate CPUs for latency-sensitive tasks like Plex or gaming VMs.

Avoid assigning shared cores between critical and less critical containers to prevent performance bottlenecks.

 

Understand Your Workload:

CPU-Intensive Tasks: Containers like binhex-qbittorrentvpn that perform heavy downloading, encryption, or compression need more CPU cores. (2-3 Cores)

Lightweight Tasks: Services like bazarr, sonarr, or radarr may not require significant CPU resources. (1 or 2 cores)

Media Servers: If Plex-Media-Server does real-time transcoding, assign more cores. (would easily eat upp all 4 on that CPU with 1 watcher...)

 

Which is why i would sugest moving plex application and runtime elase where due to resource overhead and bottleneck.

 

Edited January 3, 20251 yr by bmartino1

Is this possibly a brand new setup?  Is plex currently finding metadata?  I used to run plex on Celeron G850 and it ran fine except when plex was looking up new media data.  Have you tried unpinning your cores and letting unraid manage them? 

On 12/30/2024 at 2:23 PM, tenletters said:

changed the docker appdata / qbittorrent downloads paths to /mnt/cache instead of user to use the cache pool.

Not entirely clear what exactly you changed, but in any case, making a change won't necessarily affect files that are already on the array.

 

Attach Diagnostics to your NEXT post in this thread.

4 hours ago, Gragorg said:

Is this possibly a brand new setup?  Is plex currently finding metadata?  I used to run plex on Celeron G850 and it ran fine except when plex was looking up new media data.  Have you tried unpinning your cores and letting unraid manage them? 

It's in the off cycles. I think qbittorrent is more of the hog. Especially when it's moving files completed to the media folders. It has improved a ton since I changed my two cache drives to operate as a raid 0 instead of a 1. 

12 hours ago, bmartino1 said:

https://www.terra-master.com/global/alltos/

terra OS may have kerneal level optimizations for that hardware. unraid has a striped down version of kerenal 6 for slackware. 

in widnows therms this is akin to installing the chipset drivers. Unraid kerenal may be pulling a generic intel version which is not 3rd party optimised.
But thats a different rabbit hole and without a plugin to install driveers and config settings...
 

Based on TOS and your issues before its was also over taxed.

its not a bad cpu.
https://www.intel.com/content/www/us/en/products/sku/212322/intel-celeron-processor-n5095-4m-cache-up-to-2-90-ghz/specifications.html

 

i dont' have a good recomendation due to the number of concurent runnign docker...

 

Reserve CPUs for the Host:

Ideal with unraid: you would Leave at least 1-2 cores (preferably low-numbered ones) unpinned for Unraid's base OS and essential services.

 

Not sure of a bios or other settings that may or shold be enabled on that nas box:
 

Balance Across Cores:

You Spread workloads across physical cores to prevent overheating or throttling.

If you have hyperthreading enabled, ensure you're utilizing real cores over virtual ones for heavy tasks.
 

Isolate Critical Services:

Dedicate CPUs for latency-sensitive tasks like Plex or gaming VMs.

Avoid assigning shared cores between critical and less critical containers to prevent performance bottlenecks.

 

Understand Your Workload:

CPU-Intensive Tasks: Containers like binhex-qbittorrentvpn that perform heavy downloading, encryption, or compression need more CPU cores. (2-3 Cores)

Lightweight Tasks: Services like bazarr, sonarr, or radarr may not require significant CPU resources. (1 or 2 cores)

Media Servers: If Plex-Media-Server does real-time transcoding, assign more cores. (would easily eat upp all 4 on that CPU with 1 watcher...)

 

Which is why i would sugest moving plex application and runtime elase where due to resource overhead and bottleneck.

 

I'll check this over in a bit. Been crazy busy with work. It has improved a ton since I changed my two cache drives to operate as a raid 0 instead of a 1. I understand this gives no safety net. I set mover to run every four hours.

I do agree plex would be best on its own machine. I don't really have anything to run it on that would support transcoding. I almost never transcode, but somebody watching out of the house might here and there. 

 

Thanks for the good info. 

4 hours ago, trurl said:

Not entirely clear what exactly you changed, but in any case, making a change won't necessarily affect files that are already on the array.

 

Attach Diagnostics to your NEXT post in this thread.

For docker apps. The path for appdata defaults to /mnt/cache/appdata and I changed it to use cache instead.

But are you sure no files for appdata (and other related shares) are on the array?

 

6 hours ago, trurl said:

making a change won't necessarily affect files that are already on the array.

 

1 hour ago, trurl said:

But are you sure no files for appdata (and other related shares) are on the array?

 

 

I'm totally new to this, but I think I have it right. I might move downloads (torrents) to cache. The transfer from radarr to my media folder locks things up. Attached here the configs

Everything you have posted only tells us how you have things configured now.

 

8 hours ago, trurl said:

making a change won't necessarily affect files that are already on the array.

 

Attach Diagnostics to your NEXT post in this thread.

 

Your appdata and system shares have files on the array.

 

Nothing can move open files, so you have to disable Docker and VM Manager in Settings before these can be moved.

2 hours ago, trurl said:

Your appdata and system shares have files on the array.

 

Nothing can move open files, so you have to disable Docker and VM Manager in Settings before these can be moved.

But shouldn't they have files on the array? 

 

Under the "Shares" tab I have:

appdata Cache <-- Array

system Cache <-- Array

 

So wouldn't it copy to array then copy to cache?

 

Or do you just mean files are stuck there and not completing the move to the cache from the array? In that case, I just disable then re-enable docker/VM Manager and things should copy back and forth as normal from here on out? I had it disabled last night when I made the cach raid 0 change, but it must not have taken. 

 

I apologize for the newbish questions. I'm slowly catching on. 

You want these all off the array so Dockers/VMs will perform better and so array disks can spin down since these files are always open.

 

They can't be moved while Docker/VM Manager is enabled. So you can't enable them again until the move is complete. By default, mover runs once per day in the middle of the night. You can manually invoke Mover in Main - Array Operations.

12 minutes ago, trurl said:

You want these all off the array so Dockers/VMs will perform better and so array disks can spin down since these files are always open.

 

They can't be moved while Docker/VM Manager is enabled. So you can't enable them again until the move is complete. By default, mover runs once per day in the middle of the night. You can manually invoke Mover in Main - Array Operations.

Awesome thanks. I got that done and restarted docker. I'll check back tomorrow and see how it's running. I just got a callout for work, so I have to run, Thanks again. 

31 minutes ago, tenletters said:

I got that done

Post new diagnostics and we can take a look.

 

Or you can take a look yourself. Go to the User Shares page and click Compute... for appdata and system to see if they still have files on the array.

It's all for sure running off cache now. Thanks for the help. It struggles importing when radarr finishes, but that's alright for the time being. I think it's just how it's going to be. 

Edited January 4, 20251 yr by tenletters

Just as another update, this was needed https://trash-guides.info/File-and-Folder-Structure/How-to-set-up/

 

Sonarr/Radarr was seeing folders as different mounts. I needed to fix my folders as above so it would not copy slow and use a lot of resources. I had to follow the folder tree exactly. Exact same folder names for some reason. 

 

I'm not going into the red anymore. 

Edited January 5, 20251 yr by tenletters