FlyingTexan

Why would everything reset like this? Is this going to happen every time there's a power issue? Ok so that deleted all the dockers and now I have to go through each one and reinstall them all? Looks like it deleted my custom netowork also. What's the reasoning for all of this just because a computer lost power?

My server lost power shortly ( my fault I hit the button on the back power supply battery). Now my dockers have lost their configs. Plex, radarr, overseerr, all acting like they're being ran for the first time. Can I do something to restore things to how they were? I'd done an appdata backup but seems like it's not needed because the appdata is still there just the dockers didn't load with it.

Yesterday I installed two nvme drives. I have a 1TB for the appdata and system files and a 2TB for the server for sabnzbd and everything else. This morning I woke up to my server having crashed. I did a reboot but when I looked the nvme drive had 1.2TB of free space. I started mover which I have running every 3hrs and there were no issues. Help please 🙂 my sab and plex both lost their configurations. Tower Diagnostics 7-19.zip

I'm trying to move everything form my SSDs to my array so I can replace the drives. I've put everything to Cache: Yes. I've stopped docker. I've restarted the server. I invoke mover and it sits at zero. I'm at a loss as to why. All I really care about is getting the appdata onto the array so I can swap the drives. I've also ran new permissions but no changes. Everything sits in zeroes. Thoughts? tower-diagnostics-20230717-2203.zip

I didn’t think the array supported SSDs at all. They always said SSDs don’t support TRIM but I just bought two that do. Being able to transition from spinning runs to SSD would be amazing.

Just doing a little curious reading and trying to understand why a server board is considered to be better. https://www.newegg.com/supermicro-x12sae-10th-generation-intel-core-i9-core-i7-core-i5-core-i3-pentium-celeron-pr/p/1JW-0006-00HR4 This particular board looks to have small VRM heatsinks and a small passive cooling heatsink on the bridge. Compared to something by ASUS that's consumer grade this doesn't exactly look like it's built anymore robust. So what makes them more capable of 24/7/365 operation?

The 770 should have the newer encoder. You’re saying it’s no faster than 13th gen? 14th gen should be the same encoder as the ARC GPU

I'm not saying this will solve your issue at all but I did purchase, put together, and install a PiKVM. I was having some random issues, and I'm not home very much, so it was very hard to diagnose what it could be. Turns out my issue is a bad USB key. But what was great is that even with unraid having crashed, so I couldn't pulled any diagnostic, I was able to pull up the piKVM and see the error codes there. If you're sitting next to your server and have a monitor on it then that's great but for me it's been a life saver. Luckily I don't have any errors to show at the moment but here's a current screenshot of it.

I've been having some random USB errors. This time made it a couple weeks. I'm on a second key but have been on holiday for a couple weeks so threw in a cheap plastic $2 one. Today it dropped but luckily was able to reboot and get it running. I'll be home in a few days so just has to last me until then. I was doing more reading and picked up a USB DOM off ebay for $20. Should be here in 5 days. Will be excited to try that out.

Necro posting here but how did you setup the DOM? Did you use a desktop first then move it to your unraid? Or is it possible to install it and go into unraid and have everything transfer over?

Every article I read is about value or performance but not reliability. What's the most reliable USB key to use for unraid? Secondly, why does unraid keep checking the usb? once it boots and loads everything into memory why does it need the key to stay there? after about 13 days I just received a "Your usb key is missing" type message. Luckily I have a piKVM and on reboot it reloaded everything but I'm not home so this is a bit of an annoyance.

You’ll need a domain either way. Just pick something easy to remember and cheap then do it for 5+yrs

With Intel 14th gen around the corner will there be support issues given the igpu will be of the same type as the ARC GPUs?

I have the A380 sitting in a drawer. Honest question. Is there any quality difference between what these cards can put out vs NVENC? For all the hassle and lack of support I may just flip this card to anyone that wants it and grab a rtx 4060 (8GB) or rtx 3060 12GB. I know the 4060 has the new 8th nvenc encoder vs the 7th gen 3060 encoder but I don’t know if that’s just to add av1 support or if there an actual visual improvement.

To do this are you saying I’d have to upgrade to 6.12.2? Given my past experience with that 6.12.1 did to my system I rolled back to 6.11.5 and don’t want to leave.

When I've done diagnostics in the past it was pretty quick. Which I click on the options now, it looks like it's doing a scan of every single file I have including movies, tvshows, etc. At some point it eventually will hang and after an hour nothing will happen. Is there a means to do an abbreviated diagnostic that skips going through each file? That doesn't seem really relevant to what I'm wanting the diagnostic for. With a potential SSD or memory issue I'm hoping to pull the hardware info.

My advice is to get rid of NGINX and do a cloudflare zero trust tunnel. Removes another potential point of failure and was very easy to setup. Made my life much easier.

How do you find the performance? I don’t meet a ton of streams but the ability to transcode fast is a priority that way I can download content to iPad or MacBook.

I don’t want to take my system offline to run a memory test. Is there a means app/plugin/command to run that would let me test within the system?

Is this as native unraid or is this in a VM passthrough? I run my Plex server on my unraid natively.

Thank you for the response. If I hadn't clicked on the drive I wouldn't have seen this at all so the system wasn't alerting me to anything. So based on this you're thinking the SSD's are fine and I need to replace the memory? What happens when it detects data corruption? Is it corrected so not a huge issue or is that bad data being passed along? Is there a test or something I can do that can confirm it's a memory issue? On another note my tower diagnostics appears to have frozen. It's been running for a half hour and while showing the "busy dots" at the bottom it hasn't moved past it's current position in about 15 min. Is there a way to do a quick diagnostic and not the full scan of every file on it?

I've had the same SSDs since I started unraid in 2020. As far as I can tell my system has been running fine. It's a pure plex/learning server for now. appdata is on cache. Files are downloaded to cache then moved with mover. No hangs or anything so far. I'm working on getting the diagnostics. It's been running for a long time and I'm on hotel wifi which isn't the fastest so hoping maybe this will be enough info.

With that I could point it to a specific file to delete then tell it to restart Plex?

I’d like to setup unraid to delete the Plex codex folder then restart Plex daily at 5am. This would get rid of its codec/client comparability issue. Is there anyway to make this happen?

I've been having issues with my server not playing specific transcodes of TrueHD/DV content. There's two potential issues here. 1) The client playing plex doesn't have the same codec level as the plex server. The "easyaudioencoder" file on the server doesn't update on it's own. So you delete the codec's folder and restart plex. This syncs the codec in the server with the codec on the player. 2) There's also an inotify issue where it hits the limit. I've just updated my inotify limit which will buy more time but how do I have the server reset this number? Will only a server reboot do it? Is there an app that will auto delete a file folder? I'd like to set a time like 5am daily to have the codecs folder in plex deleted and have the app restarted. Can I automate this?