Phoenix Down

Thanks for the new docker! I'm still on the old nunofgs docker and want to migrate to your new version. However, I have a ton of plugins and customizations. What's the best way to bring them over to the new docker? Should I just copy everything over from /mnt/user/appdata/octoprint?

For sure, when I do the migration, I'll post an update in this thread.

That's what I'm using as well. I noticed the same thing - it hasn't been updated for a long time. There is another Octoprint docker in Community Apps that seems to be frequently updated. I plan to migrate over, but haven't had the time to do it. You can update all of the plugins EXCEPT for Octoprint itself. That will blow everything up, as you saw.

Pro is free for home use (right now).

My entire library is in HEVC. I use Infuse on Apple TV and have no issues playing them. I also tried the Jellyfin app on the iPhone, which also has no problem playing them. But for some reason, Jellyfin app activates transcoding even though my phone is capable of decoding HEVC in hardware. Infuse will always direct stream the video if the client has the hardware to decode the codec though (which Apple TV does).

My understanding is that if the Jellyfin app determines that you can play the video without transcoding (i.e. Direct Streaming), it will prefer that to transcoding.

I don't have an answer to your question, but how did you get that output?

Probably doesn't help you, but I don't have this issue with Infuse + Apple TV.

Yep, it's super unintuitive. You can also see if transcoding is working by going into your Admin Dashboard. Any streams will show up there, along with transcoding info, if any.

Try setting Hardware Acceleration to "VAAPI" instead of "Intel Quick Sync".

Hooray! Dashboard widget no longer blinks

Well, I'm out of ideas. I would try the Jellyfin forum as well, if you haven't already.

What format is your video file in? Have you tried a different video file to see if anything changes?

On mine, both "card0" and "renderD128" are owned by root/video. Try changing the owner and group of "renderD128" to match.

Works on my 9th gen Coffee Lake, and my understanding is that there is no change to QSV in 10th gen. What is the output when you run this in the terminal? ls -l /dev/dri

In this case, I recommend getting on their Discord channel and tag the developer there. Don’t bother with their subreddit. Tagged him in there but he never answered. Got a response pretty quickly in Discord though.

It's a display bug. It always shows that the first preset is selected, even though the preset you chose is saved correctly. You can verify this by looking at the DB table. I reported it to the developer a few months ago and he acknowledged it and said he will fix it. I'm surprised he still hasn't fixed it.

You need only look a few posts up:

I have a slightly different Gigabyte Z390 motherboard, but see if my workaround works for you. There are some caveats though, so make sure you read through my entire post, including quoted text. https://forums.unraid.net/topic/34889-dynamix-v6-plugins/?do=findComment&comment=855929

Thanks for this! I was tearing my hair out because QSV decoding was just not working. That is, until you saw your post and changed it to "VAAPI". Then it worked perfectly.

IMHO CA User Script is just a GUI for simplifying the process of setting up cronjobs. If you know what you are doing, you absolutely could just setup the cronjobs yourself.

Which is why I prefer to run the monitor process as a cronjob and not as a daemon. No need to monitor the monitor. It starts up, does its check, and finishes, all in a second or two.

Well, the easiest solution is to ask @Squid to add a delay between sending SIGTERM and SIGKILL. I generally use 8 seconds to allow processes reasonable amount of time for cleaning up. You can also have a monitoring process whose sole job would be monitor the process(es) you care about and notify you if they disappear. You might want to do this any way, in cases where your main process dies without being able to send a notification (for any reason, not just from SIGKILL). This monitor could be a daemon or a cronjob (I prefer the latter).

Power outages are rare where I live too, but it only takes one to cause data corruption. That's why I have all of my computer equipment on UPS.

What is the end goal here? Are you just wanting an email notification if your process is killed/dies? Presumably, the only time when it cannot send the notification itself is when it's terminated by a SIGKILL.