(Small).png.20e01e54adf9a0993af949e43e312763.png)
CorneliousJD
-
Posts
692 -
Joined
-
Last visited
-
Days Won
1
Content Type
Profiles
Forums
Downloads
Store
Gallery
Bug Reports
Documentation
Landing
Posts posted by CorneliousJD
-
-
Yeah, you can run it on host mode. Or give it its own IP too? Up to user discretion
-
On 3/25/2024 at 9:54 PM, Defq0n said:
The current nvidia drivers work fine with this. But be sure to read the notes when installing. You will need to stop and then start the docker service after the nvidia installer finishes. This allows docker to be able to use the gpu with the UUID of your card. Also, when configuring plex, be sure to open the advanced view in the top right corner when creating the container. In the extra parameters section, be sure to add "--runtime=nvidia" as shown in the first screenshot.
Lastly, make sure you pass the extra variables in the plex container as well, or any other container you wish to use the gpu with. These are shown in the second screenshot. The variable names are NVIDIA_DRIVER_CAPABILITIES and NVIDIA_VISIBLE_DEVICES. The last variable is where you will copy the UUID in the nvidia driver module.
If you run into any problems, just let me know.
This worked great on official container, i was missing the --runtime-nvidia part and couldn't get it working w/out that and thought i was losing my mind
-
1 minute ago, ich777 said:
Can you send me a screenshot please?
Latest should never be the open source one.
The open source option is a dedicated option and not bound to latest, production, new feature branch or a static version listed below. The version numbers might be the same but the open source is a completely separate package and should never be latest.
Just stick with latest, that should be fine.
Ok thanks - that's where my confusion was, the Latest and Open Source were the same version number, and I assumed that it was auto-selecting the open source option.
I understand how that I look at it a bit deeper (Today is the first day I've seen this plugin page option)
Sorry for the confusion, but thanks for the explanation!
-
2 minutes ago, ich777 said:
The default one.
If you are planing to use them in Docker containers for HW acceleration or LLMs not much difference at all.
Latest is fine.
Wonderful, thanks for the quick reply. I'll be using it just for HW transcoding and maybe some LLM tinkering in the future.
I'm just not clear if I need the nvidia.conf file in the modprobe.d since "latest" right now seems to be the open source driver -- it mentions workstation cards and geforce cards, but the Tesla is technically neither, and is a "Datacenter card"
I think I'm going to just bit the bullet and order the card now.
-
I've tried looking through this thread but there's nearly 160 pages now.
Can someone tell me if I install a Tesla P4 card (datacenter card) into my server, which driver should I be choosing?
I'm not particularly confident that I know the option differences here, what are the pros/cons to the production/new feature/open source drivers?
Will any of the 3 work w/ the Tesla P4?
-
Hi there, I am going to be ordering a Tesla P4 to do hardware transcoding with but I'm finding only very old info about how to get Plex to see it and use it.
I'd very much prefer to stick w/ the official container, but I'm unsure what new variables I need to add, and which specific devices I'd need to pass thru, if any? Old spaceinvaderone videos just had device UUIDs being added via variables but not actual devices being passed thru.
I'm also not sure which driver version I should be picking/using w/ a Tesla GPU with the latest nvidia driver plugin.
Is the official production one needed for Tesla GPUs (or preferred?) or should I stick w/ open source driver?
Any help is appreciated!
-
12 hours ago, JorgeB said:
I still think it must be one of the containers, but possibly doesn't do the issue immediatly.
I wanted to make sure to post here for anyone else who finds this later, but i think this may be resolved.
So there was definitely an issue somewhere, my MariaDB got corrected and that may have been the culprit in the end?
I re-created docker folder as docker.img and re-added containers and they all added fine ,server was stable, but upon rebooting w/ docker autostarts on, it went to 100% again.
I was patient this time and let it sit and it stabilized after 10-15 mins or so.
I know I have a lot of containers and this is likely to be epxected.
Whatever was initially going on though, the server wasn't stabilizing and it was sitting at 100% CPU for hours, so perhaps a combination of fixing my MariaDB (restored from backup) and/or re-creating docker folder as docker.img ended up resolving everyting.
I'm at about 24 hours of stability w/out hitting high CPU now.
Thanks JorgeB for the help. (Again!)
-
1
-
-
Nearly every single container is telling me there's a shared mapping, is this normal? It doesn't seem right.
Lidarr's appdata for example isn't mapped anywhere else
-
6 hours ago, JorgeB said:
If you are still using a docker folder I would also try a docker image instead, folder has been known to sometimes cause weird issues.
I've completed everything above, during recreating my docker.img file I loaded everything back from previous apps in CA and they all ran fine, sitting at like 30% CPU usage.
I setup my autostartup on docker containers I wanted auto-starting and rebooted and I'm back at 100% CPU again...
I'm at a total and utter loss here.
Is there anything I can look into? The server becomes unusable after it sits like this for a while.
-
5 minutes ago, JorgeB said:
If you are still using a docker folder I would also try a docker image instead, folder has been known to sometimes cause weird issues.
Interesting, I'll move back to a docker.img file -- i figured without that layer of the img file that i might see better performance w/ just a folder instead, plus never hitting limits (as you can see I have a LOT of containers running)
I'll let the current btrfs balance finish first then kill off the docker folder and re-create a docker.img instead in /mnt/cache/system/docker/docker.img and re-fire up containers and see what happens from there.
-
Just now, JorgeB said:
I suspect one of the containers, but difficult to confirm without trying one by one.
I left docker running and turned them all off one by one and watched my CPU but no changes, I ended up having every single container in a stopped state and CPU was still at 100% with dockerd being the culprit still. Very strange to me.
I'm running a btrfs balance right now and CPU is back to 100% usage but it was shooting up before I started that even.
Currently working so it's difficult to dedicate a ton of time to this during the work day but I'm starting to now regret my decision to go NVMe because everything was absoultely fine before this.
I'm wondering if its somehow related?
-
I left the server on overnight and woke up to much lower CPU usage, but I'm still wondering what could have caused this - the server became so sluggish it took an extremely long time to do anything.
Also btw the top command showed 0.0 wa (IOWait) during all of this, forgot to include that screenshot.
-
I'm at my witts end here on this one, totally at a loss.
Been at this for hours tonight and I think I need to admit I need some help here.
I replaced my cache drives w/ new ones, but found out that they didn't support deterministic trim, so I swapped them out for a Asus Hyper M.2 v2 PCIe 3.0 card that holds 4x M.2 NVMe SSDs, and did PCIe bifurcation to 4x/4x/4x/4x and that worked great.
I swapped the NVMe drives in to the cache pool one at a time.
It took about 12 hours for each drive to rebuild.
I did notice my CPU pushing 100% usage pretty much the entire time, but I didn't think much of it because it was rebuilding the cache drives.
At this time I also swapped everything from using /mnt/user/appdata/ to /mnt/cache/appdata/ to eek out some extra speed.
I changed every container touching appdata to this and also my docker settings pointing to
/mnt/cachce/system/docker (using folder, not docker.img)
and /mnt/cahce/appdata/
I rebooted the server for good measure (glad I did because I found it tried to boot from NVMe and not my flash anymore, whoops!)
and then I realized my MariaDB was corrupted and NextCloud stopped working.
Weird... okay. moving on, I saw my CPU was stuck at 100% usage when Docker was running.
I stopped ALL containers one by one and CPU was still at 100%, which didn't make much sense, all containers in stopped state still had this problem.
At a loss, I deleted my docker directory (I'm using a directory, not an img)
Everything started coming back up fine after re-adding them from CA, with the exception of course of Nextcloud/MariaDB.
Thought I was in the clear and it was a corrupt docker folder at first, but while troubleshooting the MariaDB/Nextcloud issue, I saw my CPU shoot back up to 100% again. Very weird. And that's where I'm stuck now...
I ended up removing MariaDB for now entirely and deleting its img but still no luck.
As you can see I'm stuck at 100%, but overall usage per container is very low.
Diagnostics and screenshots attached below.
-
2 minutes ago, JorgeB said:
Those boards have plenty of lanes, with 2 CPUs you can fully use all four x16 slots, so as long as one of those is available, and it can be bifurcated it's not a problem.
Excellent, thank you.
And thanks for your help on my SSD trim question too. This is a direct result of your help there.
Much appreciated.
I'll be getting this card and some NVMe drives for cache then.
-
1
-
-
Hi everyone,
I have the following setup
Motherboard: SuperMicro X9DRi-LN4F+
https://www.supermicro.com/manuals/motherboard/C606_602/MNL-1258.pdf
CPU: 2x Intel Xeon E5-2650 v2
SAS Card: LSI 9211-8i HBA
https://docs.broadcom.com/doc/12353333
I am wondering if I buy this ASUS Hyper V2 NVMe to PCIe adapter that I'll have enough lanes to run it in 4x/4x/4x/4x bifurcation and not risk taking any lanes away from the LSI SAS card?
https://www.asus.com/us/motherboards-components/motherboards/accessories/hyper-m-2-x16-card-v2/
I have already updated the BIOS to v3.4 so bifurcation is enabled/supported on my motherboard.
I am seeing that each CPU has 40 lanes available and I have two CPUs so I should have plenty of lanes to spare, however I'm not super familiar with the layout of the PCIe lanes and which one links to which CPU etc.
Can anyone help shed some light on this for me?
How many total lanes do I get between my CPUs and my motherboard?
I couldn't really find info on how many lanes are available on the mobo side.
I really just want to make sure I'm not sacrificing performance or ripping lanes away from the LSI card to put in the M.2 card.
Thanks in advance!
-
17 minutes ago, JorgeB said:
Are they connected to an LSI? They only support trim for devices with deterministic trim support.
They are indeed.
Is not having trim an issue? I'm admittedly not very familiar with its purpose.
If so, in your opinion should I get a PCIe card that holds 2x M.2 drives and return the 2.5s and get M.2s instead so they are internal and on a different controller?
-
fstrim: /mnt/cache: FITRIM ioctl failed: Remote I/O error
I had some older 2.5" Samsung EVO SSDs that had been starting to give me problems so I replaced them in Inland (MIcroCenter) enterprise SSDs just the other day.
Ever since swapping them both out, I am getting the above error emailed to me on a daily basis.
The new drives are connected to the same ports on the backplane and everything and I've never seen these errors before with the old drives.
Any help is appreciated!
-
20 minutes ago, variants said:
Sounds correct, you have to edit the docker to add ports to the docker, you can add a bunch of ports one time to not have to stop it every time you add a new game server to Amp like 8200:8300 tcp and udp
Was just going to suggest adding ports ahead of time that you anticipate needing.
If you want to use default ports for games then yes, you need to edit the container every time to pass the correct ports through.
If you're hosting *that* many games servers where AMP within Docker like this is problematic then perhaps a containerized AMP isn't the best solution for you?
-
7 hours ago, wgstarks said:
Quote from the LSIO docker page
If you search you’ll probably find a few discussions about the change but the basic reason LSIO is giving is that maintaining the Unifi Controller docker has been too much trouble. The complaint that I’m seeing most often regarding the new Unifi Network Application docker is that you also need to run a separate MongoDB docker. It’s not an all-in-one image. I did a little how-to post in the Controller support thread for switching to UNA. It’s really not very complicated to setup but it does add an extra point of failure.Ah thank you, I was worried for a moment this was some change coming down from Ubiquiti, glad to see this project is being taken up to have an all-in-one container then. Keeps it simple for most. Thanks
-
1
-
-
12 hours ago, PeteAsking said:
Its not being discontinued like the linuxserver.io one
Oh really? That's interesting, is that coming from Ubiquiti that's discontinuing the controller for linux or is it linuxserver discontinuing theirs?
I hadn't heard of this so now I'm extra curious, as I know quite a few users who are self-hosting their controllers, but myself am not.
-
I'm not currently clear what this is for vs the other UniFi network/controller containers? What makes this different?
-
1 minute ago, CatDuck said:
Am I reluctant to make another template because of duplicates. I understand I can absolutely do what you mentioned.
Think if a few months go by and the fork is being maintained and other isn't, that we revisit makeing the change?
I'm open to that if there's still no movement on original project.
You can edit your existing container to point it to the new work though if you want to give it a run for the time being still
Without making a new template in CA.
-
1 hour ago, CatDuck said:
Hey @CorneliousJD
Do you mind updating the repo to a new work thats not abandoned?
https://github.com/Cesura/pastey/pull/33#issuecomment-1777021527
You can just update the dockerhub repo to arcscloud/pastey:latest and that should pull their copy instead.
I'm reluctant to adjust the template in the community apps section since arcscloud's work is just a fork of Cesura's original work.
-
3 minutes ago, BennyD said:
Can we just get you to completely drop this project so that the actual Dev for AMP will maintain this?
WILL they? They don't supply a docker container.
Why would they maintain a third party container?... this doesn't make sense.
I also didn't develop this container, I simply made the template to run it on unRAID.
I cannot assist everyone with issues they face with AMP.
I have a full life outside of this container.
The container itself runs. AMP runs. I host servers with it. It works. It is not beginner friendly, but it works.
[SUPPORT] JoplinApp - CorneliousJD Repo
in Docker Containers
Posted
Yes you guys just need the Joplin server (there USED to be an app in CA for it but it's been removed) or just run it serverless since it uses flat files.
no actual need for a server. the idea behind this (not developed by me) was to just have a way to view your notes in a web browser was all.
This is not a server side software -- it just ran the app w/ a VNC window so you could view it.