MowMdown

You can resume the data-rebuild, disks are mounted.

That sounds like a bad cable or a bad NIC port somewhere in the chain. This wont be an unraid issue.

I have a DXP2800 that I can fully saturate at least 1Gbps. Which I’ve done for a full duration of 9TB during a data backup session. This was strait to a btrfs raid1/mirror pool of 2x 12TB HDDs Unraid to unraid via restic over sftp The one thing that can dramatically cut speeds is unraids FUSE layer, if you can bypass FUSE with an exclusive share it can greatly speed up copying. With that said FUSE seems to only be affected by many small files being read/wrote. A single large file shouldn’t really be an issue.

Run an iPerf3 test from your windows pc to the server and confirm it’s not the networking.

PCIe devices like GPUs downgrade their gen when not in use.

That's what I posted earlier about, see below: However with that said, you shouldnt be using disk-pool names and instead be using /mnt/user/appdata/... for your paths to appdata

You have to be careful even with PSUs from the same manufacturer and of the same model. Unless the manufacturer says you can use the same cables, assume it's not safe.

Try this

Please post system diagnostics

It doesn't matter what the reasoning is, posting the prompted result without any self-formatting and revision just tells me you took no effort to formulate a response. Using LLM for assistance is fine, just dont post the raw result because if that's the answer I wanted, I would have just done that myself. LLMs are not accurate and their output needs refinement and correcting multiple times.

DOCP profile is an overclock FYI, you might have heard people refer to intel's equivalent of "XMP" or the newer AMD "EXPO" These are all overclocked memory profiles. You would need to disable DOCP.

There is a difference between using AI to act as a translator and using AI to generate an answer to a question. If the person proposing the question wanted an answer from a clanker, they would have just asked the clanker. Translation tools have existed long before AI and are perfectly capable converting text between languages. AI isnt needed for this. If you want to bounce ideas off of an LLM, you should still format said thoughts into your own words and feelings and not just lazily post the LLM generated text as a formal response to a question. Lazy and wastes everyone's time. This response was NOT formatted with AI assistance nor was AI used in the process.

I have not confirmed this with any staff but... I believe limetech changed how the array runs, and yes when the license expires the array is stopped to prevent it from running indefinitely on an expired timer. Ive noticed this behavior on trials and when the USB is removed while only acting as a license and not a boot device.

Memtest should be ran minimum 24 hours

Depending on which version of unraid you updated from, the linux kernel had a jump in versions which could have introduced driver changes to the network driver or USB drivers. If blacklisting the sleep function makes it work then yes that is the correct solution. Just keep in mind this behavior can change when when unraid updates and uses a newer linux kernel.

Regardless of how the system is turned off, when mains power is restored the bios should still turn on the system like it currently does when it detects power. I personally prefer using the NUT plugin to manage the UPS and system. Mains power is lost, UPS switches to battery. Unraid runs 10 minutes on battery before a shutdown is initiated by UPS signal. Upon system shutdown the UPS is also powered off after Unraid is powered off. Mains power is restored and system boots up, UPS comes online. Ideally this is how it should function. You will need to research the UPS supports this.

Just make sure you configure notifications in the settings to get disk alerts

Make an entire copy of the existing USB and save it somewhere accessible outside of unraid. Flash a new install of unraid to the new usb. Copy the entire “config” folder from the backup you made and copy it to the new USB. Alternatively you can just redo the exiting USB with a new install after taking the backup to really be sure the existing USB is truly no good.

I mean, if you get an actual JBOD, you would get an LSI card that ends in "e" -> LSI-9400-24e which then has external connections to a JBOD. Alternatively you get a server rack style case that holds as many drives as needed which you would get an LSI-9400-16i "i" mean internal connection. these are just examples, you would need to find a card that is appropriate for your system hardware and needs as there are many makes/models.

Changes in the linux kernel

USB appears to be goobered. back it up, install a clean fresh install onto it, restore the whole config folder from your backup and try booting again

What likely happened is there was many docker containers using RAM in high quantities which resulted in 99% RAM use on the system when that error was generated. Did unraid crash? I see there was a reboot since the logs start on July 19th but I cant tell if that was due to a hard shutdown or you just rebooting it.

Based on what I saw, I would have guessed that one in particular was being problematic. I just wasn't 100% sure. This is more of a general statement, enabling settings that is outside of stock unraid carries the potential to cause instability. This applies to anything the BIOS manufacturer didnt ship stock too. Going back to the defaults is just a step of troubleshooting to make sure the changes you made didnt affect it in some way/shape/form.

I dont want to give a false impression but I've been running a Ryzen 7 5800X in my server for years now and have not had these issues. With that said Im running a B550 motherboard which is a newer chipset and different bios which could also play a role in your issues. I suspect it's the ancient 1800X CPU but I would hate for you to go spend money on something that doesn't fix your issue. 1st gen ryzen was pretty hit/miss

Another thing is you are using plugins and 3rd party packages which alter the system in ways which might be unintended such as powertop and autotweak.These could cause system instability.