Thanks to all for sharing your efforts here, I’ve just bit the bullet and ordered a Gen 11 i5 Nuc myself after finding a deal too good to pass up. I notice the same CPU as your unit @Atlantis5 - Hope it’s still going well…
I’ll share my results when it arrives and I’ve had a crack at getting Plex up and running. Cheers! 
FYI, I’ve managed to transcode a few H.264 files to H.265 (CBR, VBR and quality-based) for personal purposes with ffmpeg and QSV plugin. Package is v4.3.2 from Ubuntu Hirsute.
Strangely, I’ve transcoded dozen of files, and got ZERO crash… The transcoding performance is huge! I’m confused, I don’t understand what causes Plex crashes…
Oh dear… I was actually planning on buying one of those new NUCs, but this looks like too much trouble as it is. Is this problem occurring with Windows 10 as well and should I go for a Frost Canyon instead?
Edit: Seems to have been answered, so would like more confirmation before I waste my money 
Thought I would add my experiences to the thread …
Just purchased a new Nuc11 i5, and knowing Ubuntu 20.04 didn’t have the drivers for the GPU went straight to Ubuntu Desktop 21.04 (also wanted to play with X11 - that project now abandoned). I run plex and other apps on docker.
After some initial successful tests transcoding content (including 4K) I saw that the plex container was crashing. After this happened several times also found that while my ssh session appeared ok, the ubuntu desktop had also crashed and all I had was a single blinking cursor on the console and needed a power cycle to reboot the server. Happy days.
Next I thought I’d try going back to a clean install of Ubuntu 20.04 and upgrading to 20.10 to pickup the GPU driver support. Several rounds of testing later proved that same issue was occurring, and checking the kern.log file was seeing reassuring messages such as GPU hang. Cool.
It was at this point I came across this thread and experienced both the joy of knowing I was not alone, and the despair that this doesn’t seem to have a reliable solution yet. As there were hints pointing to NUC bios version may be an issue I bit the bullet and updates to latest bios for my NUC (40).
Everything went downhill from here.
I then started seeing on console a bunch of errors complaining about I/O on my m2 drive. Messages such as:
blk_update_request: I/O error, dev nvmn0n1 …
Buffer I/O error on dev dm-0 … lost sync page write
Aborting journal on device dm-0-8
JDB2: Error -5 detected …
… Unable to read boot sector to mark fs as dirty
Long story short after performing an in-place upgrade to 21.04 and also several clean installs of ubuntu wiping the entire disk these errors will occur not long after booting. The server will not respond to reboot or shutdown commands, and once again power cycle is required. Note again this is on a clean install of Ubuntu server with no apps deployed.
Attempts to downgrade the bios have failed which is nice.
Have also tried re-seeding the m2 drive and disabling secure boot in bios. Nothing works, and I now have an unusable NUC and need to head down the warranty path.
I would hold off buying a Nuc11 for now. Would be interested to know if anyone else has experienced the disk issues though.
It seems that there’s not that much support from Intel for their 11th gen CPU. It has been a nightmare for me as well.
I have a 1135G7 and I always faced problems on this CPU while Plex runs fine on a J4105 …
Back on 1st July you were given a possible solution using Arch from Atlantis5. Did it work for you?
My results from the things that I have tested align with the experience of you and geeooff. So if there’s a possible workaround, I’d be keen to hear about it.
Thanks,
JT
@DarthBJW my friend, take some solace in the form of the immortal words of Michael Jackson:- You are not alone, I am here with you… And hopefully I might be able to help you. (That last bit’s from me, not Michael)
Is your m2 SSD, by chance, an NVMe drive?
I have been having similar issues with the 1135G7 i5 Nuc 11. The plan was to follow @Atlantis5’s path and run Arch Linux to experience the glory that is Plex transcoding with Quicksync on Linux, but alas, I didn’t even get to the Plex installation before the system/NVMe drive started freezing and locking up with I/O errors and EXT4 fs errors, among others. I noted the issues only began from 2nd boot onwards each time I reinstalled linux, but from every boot onwards the issues would recur between 5 and 60 minutes without relent.
Ironically, Windows 10 had no issue with the same drive, and all tests confirmed there’s no obvious issue with my NVMe drive. I’m somewhat ashamed to admit until I stumbled across this little nugget below, I almost hit the F it button and went with Windows, but I held strong and resisted the powerful pull of the dark side.
According to THIS bug report, there are known issues on Linux that can affect some NVMe drives by way of lock up/freezing, potentially related to how NVMe drives are managed by power saving features. Some drive manufacturers have released firmware updates, so check for those, and there have been some kernel patches to address the issue, however even after the workaround below I note some users mention they still experience the issue very rarely, but it’s progress.
The workaround I have found stability with for the last 24 hours and counting (So far…) which I found thanks to some of the users comments in the bug link above, is below:
Find your Grub file, and using a text editor with SU privileges, add the following line to the end of your GRUB_CMDLINE_LINUX=””
nvme_core.default_ps_max_latency_us=0
I also added the below line, which may not be necessary but I’ve had no issues since I made the changes.
pcie_aspm=performance
More detailed instructions than mine can be found HERE thanks to Berk.
I suppose if all else fails, one could try a non-NVMe SSD drive, however that would slightly diminish one of my goals which was to drastically improve Plex metadata performance.
Apologies if this is somewhat unrelated to Nuc 11 transcoding on Plex, but as another user mentioned an m2 issue I thought I would offer my findings which helped.
Now, onto Plex and transcoding! …Once I have some more confidence in stability…
Hey @Samwise41 , good to hear I’m not the only one; while still sorry you are experiencing this too
Indeed yes, I am using an NVME drive and since my last post have been able to confirm:
I have a Kingston NV1 NVME drive, and while I had windows loaded up I checked with their SSD Manager and confirmed firmware is up to date.
Interestingly, when I did some testing with Linux Mint the issue didn’t occur. I gave it about an hour (which historically is ample time for the NVME to go cactus) and during that time had the epiphany that Mint is Ubuntu derived, and as Mint is a desktop flavour it doesn’t create LVM partition by default …
So another wipe and replace and am back with Ubuntu Server 21.04, but this time I unchecked LVM during the install sequence. It’s been 30 minutes so far and no disk issues have occurred so I am cautiously optimistic.
EDIT after 1 hour - Negative ghostwriter, the pattern is full 
Would be good to know from you what you currently have for bios version, linux distro, NVME model, and whether you are using LVM or not.
I’m sorry to hear that @DarthBJW, I was hoping you might have a win. I suppose running cat /sys/module/nvme_core/parameters/default_ps_max_latency_us returns a value of 0 right?
Interestingly, I’m using the same NVMe drive, the Kingston NV1 in 500GB form (SNVS/500G), yeah there’s no firmware updates for this yet, it may be worth considering raising cases with Kingston for testing if we can’t find a workable solution.
I’ve been stable since my last post , hope you’re having some luck!
I’m pretty happy with things so far, so I’m going to go forward with installing the intel driver and Plex for some transcode testing
I haven’t had the energy to try yet!
I didn’t have a lot of faith in it been resolved, the kernel versions are the same and I’d be very impressed if that one case was working.
I was more interested that @geeooff as able to transcode using the cli outside of plex, which kinda suggests there would be a way for plex to fix this on their side.
@ChuckPa Did you notice that @geeooff had been able to run tests outside of plex for transcoding that didn’t fail.
This was of great interest . First bit of information that I feel @ChuckPa could take to an engineering team or at least the start of something to understand the difference.
May need to know the commands you were running @geeooff
After a lengthy detour trying (and ultimately failing) to get the NVME drive working, I’ve ditched that and put Fedora onto an m2 SATA.
Still cant transcode in hardware. Despite what appeared to be a promising start, still see errors appear about GPU hang and ultimately plex will crash. Fedora seems to handle the aftermath better than Ubuntu did, as at least after the restart I could stream something from plex.
$ sudo inxi -G
Graphics: Device-1: Intel TigerLake-LP GT2 [Iris Xe Graphics] driver: i915 v: kernel
Display: server: No display server data found. Headless machine? tty: 238x66
Message: Advanced graphics data unavailable in console for root.
$ uname -a
Linux palpatine 5.12.13-300.fc34.x86_64 #1 SMP Wed Jun 23 16:18:11 UTC 2021 x86_64 x86_64 x86_64 GNU/Linux
@ChuckPa is there any further data I could grab which would be useful to you guys at this point?
@DarthBJW @ChuckPa isn’t directly working on this as so far the issue has been classed as a kernel issue. Meaning the fix will come from outside of plex.
However based on @geeooff last post, it’s extremely interesting that he’s able to transcode files using the QSV encoder without any of the issues Plex has. I’d hoped this would be a data point that could be passed back to Plex’s engineering department to understand if something should be done differently for these chips.
That been said, there are already workarounds for the desktop variant of Xe that didn’t require any changes of plex. So it’s only a slight hope.
To be completely fair with Plex standard behavior, I will retry to encode to H.264 (CBR mode).
FYI, here was my command to transcode H.264 to H.265 (CQP mode) using QuickSync:
ffmpeg -y -hwaccel qsv -c:v h264_qsv -i "$input" -map_metadata 0 -c:a copy -c:v hevc_qsv -load_plugin hevc_hw -preset slow -q:v $quality -look_ahead 1 -movflags +faststart "$output"
I’m struggling to transcode @ChuckPa’s H.265 HDR sample to H.264 (i’m missing the zscale filter to convert to SDR), so I’ve tested H.264/8-bit to H.264/8-bit transcoding with one of my files.
No problem at all, transcoding at >17x speed (1.5M constant bitrate).
ffmpeg -y -hwaccel qsv -c:v h264_qsv -i "$input" -map_metadata 0 -c:a copy -c:v h264_qsv -b:v $bitrate -load_plugin h264_hw -preset slow "$output"
No issue so far with an WD_Black SN850 (model WDS500G1X0E-00AFY0).
ALL,
I’ve been trying to follow along as best I can here.
Not sure if anyone is aware but I’ve been **** deep in aligators with the Synology DSM 7 Release, Release hot fix, Release into the wild 
events.
Been kinda crazy over there just trying to get some people to read the documentation 
We did get a NUC11 in the lab but, because of all the chaos around me, I’ve not had time to work with it.
I DID get the Debian packaging work completed so I can correctly detect 10th Gen and 11Gen and all future X-Gen (or was that “Gen X” ? 
) CPUs.
My plan is to get the packaging through QA this coming week.
When out to you, the preinstall phase will perform a correct inventory of the CPU and set the appropriate flags for package dependencies.
For your consideration, please review when you have an opportunity .
This is the updated preinstall module.
It will write to /tmp/plexinstaller.log when complete.
As always, run as root
preinstall.tar (40 KB)
Looks good on my i7-10700K
sudo ./preinst.sh
PlexMediaServer install: Pre-installation Validation.
PlexMediaServer install: Pre-installation Validation complete.
# Plex Media Server installation configuration info: Fri 02 Jul 2021 04:48:09 PM EDT
Init=0
Systemd=1
Custom=0
LinuxContainer=0
NewInstall=0
HaveOverride=0
OverrideFile=""
PlexUser="plex"
PlexGroup="plex"
VideoGroup="render"
AppSuppDir="/var/lib/plexmediaserver/Library/Application Support"
PlexTempDir="/dev/shm"
LangEncoding="en_US.UTF-8"
ExistingVersion=12303
HaveHardware=1
HaveTranscoderPref=1
NeedUser=0
NeedGroup=0
NeedVideo=0
Verbose=1
Running=1
Errors=0
Warnings=0
HaveIntel=1
HaveIntelCore=1
HaveIntelXeon=0
Havei915=1
HaveNvidia=0
Processor=Intel(R) Core(TM) i7-10700K CPU @ 3.80GHz
CPUSKU=10700
NeedOpenCL=1
NeedBeignet=0
NeedIntelCompute=1
HaveIntelGmmlib=1
HaveIntelIGCCore=1
HaveIntelIGCOpenCL=1
HaveIntelOpenCL=1
HaveIntelOCLoc=1
HaveIntelZeroGPU=1
HaveIntelCompute=1
HaveBeignet=0
HaveOpenCL=1
Platform="System Product Name"
Processor="Intel(R) Core(TM) i7-10700K CPU @ 3.80GHz"
Distro="Debian GNU/Linux 11 (bullseye)"
Kernel="Linux Saturn 5.10.0-7-amd64 #1 SMP Debian 5.10.40-1 (2021-05-28) x86_64 GNU/Linux"
Memory="31936 3131 20242"
Also I suggested in another thread, but any of you guys with NUC11’s try out Debian 11? Just curious, it may solve you issues. Debian is well known for its stability.
Here is the net-ISO
Thank you. Also, thank you for catching that typo which I looked right at but did not see 
ALL,
I’ve been trying to follow along as best I can here.
Not sure if anyone is aware but I’ve been **** deep in aligators with the Synology DSM 7 Release, Release hot fix, Release into the wild events.
Been kinda crazy over there just trying to get some people to read the documentation
We did get a NUC11 in the lab but, because of all the chaos around me, I’ve not had time to work with it.
I DID get the Debian packaging work completed so I can correctly detect 10th Gen and 11Gen and all future X-Gen (or was that “Gen X” ? ) CPUs.
My plan is to get the packaging through QA this coming week.
When out to you, it’s do what it’s supposed to do for inventory & reporting.
