paperblankets

I'm on a temporary motherboard/cpu with fewer PCI-e slots while my motherboard is RMAed, I suspect quite a few plugins are angry about hardware changes (In particular a GPU is missing so it makes sense the plugin is unhappy I guess).

Fresh diags after reboot ms-diagnostics-20240403-1053.zip `/mnt/user` is still linked, and the server is not manifesting errors at the moment. I'm mostly posting the diags in case they are useful for resolving the root issue. I will follow up if I notice any more issues, let me know if I can provide any useful info.

Here are some fresh diagnostics. I just got this error for the first time (that I'm aware of). ms-diagnostics-20240403-1014.zip

I removed the old user scripts. I also found weird behavior in the unraid frigate template, setting `NVIDIA_VISIBLE_DEVICES` is not respected by the image, but using `CUDA_VISIBLE_DEVICES` works as expected.

Is there a way I can re-prioritize the gpu order in nvidia-smi? ~# nvidia-smi Mon Mar 4 07:50:18 2024 +-----------------------------------------------------------------------------------------+ | NVIDIA-SMI 550.54.14 Driver Version: 550.54.14 CUDA Version: 12.4 | |-----------------------------------------+------------------------+----------------------+ | GPU Name Persistence-M | Bus-Id Disp.A | Volatile Uncorr. ECC | | Fan Temp Perf Pwr:Usage/Cap | Memory-Usage | GPU-Util Compute M. | | | | MIG M. | |=========================================+========================+======================| | 0 NVIDIA GeForce RTX 3090 On | 00000000:01:00.0 Off | N/A | | 0% 37C P2 96W / 350W | 1164MiB / 24576MiB | 2% Default | | | | N/A | +-----------------------------------------+------------------------+----------------------+ | 1 NVIDIA RTX A5000 On | 00000000:82:00.0 Off | Off | | 30% 33C P8 6W / 230W | 2435MiB / 24564MiB | 0% Default | | | | N/A | +-----------------------------------------+------------------------+----------------------+ What I am actually trying to do: I would like to use the A5000 before the 3090 for containers (but I want to be able to use the 3090 when I need to) I am using this user script to bring power states down hourly, but what I really need is for my containers to prefer the A5000. Right now I pass through NVIDIA_VISIBLE_DEVICES-ALL to each container, if I pass through a GUID list with the A5000 first would that impact priority? Thanks for letting me pick your brains.

If this is the first time setting up Plex: I think your container was restarted (or otherwise interrupted) during it's first run, and the database failed to initialize properly. Deleting /config (or pointing to a new directory) should allow Plex to initialize SQL properly. Alternatively if /config is mounted somewhere strange you may need to set the mount to a different type, or it could be a permissions issue.

It has been added there as well

During backup I get a version of this error for a few containers: ... [24.02.2024 03:32:23][❌][unmanic] '/tmp/unmanic' does NOT exist! Please check your mappings! Skipping it for now. ... But I have that, and the other failing directories in my global exclusion list. Is there a config tweak I should be doing to ignore these tmp paths? Thanks for maintaining and working on the backup plugin!

I'm not sure if this will help some of you with APC 3000 hardware. After enabling ModBus (On the UPS, and in settings) I HAD to physically disconnect the USB cable before nominal power rendered. I bet I wasted hours trying permutations of things, and in the end that's all I had to do.

To add to ConnerVTs post above, you can also configure an application on zero trust to only be accessible on your cloudflare gateway (vpn). https://community.cloudflare.com/t/bypass-login-page-when-warp-client-active/376061/11 I can't speak to the overall security or risk of misconfiguration in zerotrust, but that would give you fully 'tunneled' communication.

Container request for a vpn version of tubesync of https://github.com/meeb/tubesync/blob/main/Dockerfile

I have ruled out this being DNS related (at least internal DNS). Verbose logging it is.

Based on that I think waiting for a stable 3.0 sounds reasonable even if that's a month out. I can check for a more verbose error option in the container (I have a vague memory of a path flag that increases the verbosity of some of the service logging). I could also go to static IPs to rule out DNS as well. That's maybe where I should start.

Hi @KluthR! Thanks for your work on the new plugin. I see a v3 is coming soon and could wait for that if thats best, or test it if you think what I'm about to bring up might already be addressed. I have some containers that fail to restart after backup. Frigate is one this happens with consistently. I believe this happens because of the extra options this container is started with. The container starts with Extra Parameters: `--runtime=nvidia --shm-size=5G`, which, by itself is not to crazy. But it also has a coral AI/TPU (device) passed in. The weird thing, is the backup thinks the container starts successfully from what I see. ... [08.04.2023 03:00:13] Stopping frigate-nvidia... done! (took 4 seconds) ... [08.04.2023 04:08:32] Starting frigate-nvidia... (try #1) done! but the container just stops right after starting (In this case, I started it myself at 11:29): [2023-04-08 04:08:36] frigate.app INFO : Starting Frigate (0.10.1-83481af) Starting migrations [2023-04-08 04:08:36] peewee_migrate INFO : Starting migrations There is nothing to migrate [2023-04-08 04:08:36] peewee_migrate INFO : There is nothing to migrate [2023-04-08 11:29:36] frigate.app INFO : Starting Frigate (0.10.1-83481af) 23-04-08 11:29:36] peewee_migrate INFO : There is nothing to migrate [2023-04-08 11:29:36] frigate.app INFO : Output process started: 227 [2023-04-08 11:29:36] ws4py INFO : Using epoll ... I really don't know why this is happening, but thought I would share some of the weird in case something jumps out, or someone else is experiencing the same issues. Maybe causes: My DNS server (a container) might be down, maybe frigate is silently erroring trying to access cameras. If it is that the DNS server container has not started yet, is there a quick solution to this? Like a pre start script that starts my DNS server container. I have not tried that since I thought the backup script would error trying to start the DNS server when it was already started. Thanks for any thoughts or ideas, and the work to improve the backup plugin.