smegger68

Since my original post I have added a Quadro P400 to work with Jellyfin. I won't be using more than one stream so this is more than adequate. I have added the Dell IPMI fan control docker to my config and that nicely sorts out any noise issues and allows the fans to ramp up if anything gets hot. I could still put way more powerful CPUs and add a ton of memory in there so I can see this box lasting me for a few years yet.

I seem to be unable to log into the admin account on Tonido. Any idea how I can recover or reset the password?

No. You will have to either add the data to the array by copying it and then add the drive to the array or run the drive as an unassigned device, which of course means it has no parity protection.

Here is my tale of acquiring and updating an old 12th gen Dell T320 tower server to be a rather lovely unRaid box. I've had unRaid servers for 11 years at this point, upgrading the hardware along the way. This is still the original upgraded install from all the way back. I've been through Intel Atoms, a couple of Athlons and for the last few years an Opteron 6380 based system. This was actually quite good apart from the power usage being way too high for my liking, especially at idle. Even with all the drives spun down and nothing much going on it would still juice 110w from the wall. I don't know if you've seen the price of electricity recently but... yeah. Time for a change! I managed to pick up a T320 tower from an acquaintance for £100. It was working fine but was fairly low spec. Ok not a problem, parts for these old systems are cheap on the electric bay. What I had bought turned out to have a Xeon E5-2407 4 core 4 thread cpu from the Sandy Bridge EN family. 80 watts and not exactly stellar performance! An E5-2430L V2 was procured for £25 on the big auction site. This is a 6 core 12 thread chip which will turbo up to 3 Ghtz while only using 60 watts. Ivy Bridge EN was quite a step up My mate informed me it only came with 4GB of RAM. I have a ton of DDR3 server memory laying about so wasn't worried. When I opened her up though I found 3 sticks in there? Turned out 2 of them were not seated at all. After plugging them in properly - yep, 12 GB. This is plenty for my use case and I can always bung in more sticks if I want to. The Perc H710 & 8 drive cages were all present and correct so I began the upgrade with flashing the HBA to IT mode with the appropriate LSI firmware. There are ready guides and videos out there to explain how to do this. Once it was done I swapped my drives over to the box and booted unRaid for the first time. And it worked first time too! Honestly I could have stopped there but I wanted to do that cpu upgrade and I knew this would entail a BIOS update as the ancient image on the box had never been updated. The old BIOS has no microcode for Ivy Bridge CPUs. This is where the first issues were encountered... Dell include a middleware on these servers called the Lifecycle Controller. It's sort of a catch all subsystem that allows you to modify the settings for BIOS, RAID & the DRAC etc and also allows you to do remote updates direct from Dell as it will connect to the internet and download whatever is needed. Except... Dell changed it in subsequent revisions. It used to connect to an FTP server at Dell but they got rid of that and nowadays it needs to connect to an HTTPS box. There is NO provision to do this in the earlier LCCs. So manual updating is the only way unfortunately. There are a few options and all of them are worrisome and painful, but the best way is to use an application from Dell called the DRM - Dell Repository Manager. It's a clunky old bit of code with little documentation but with a little trial and error I managed to get it to do what It's supposed to do - which is make a bootable ISO image containing the updated LCC, BIOS and DRAC. So I booted up with this on a USB drive courtesy of Rufus and it actually worked OK. Not without drama though... it got to a point and stopped and wanted an 'encrypted repository passphrase'. This caused a bit of head scratching as I hadn't entered one. After a couple of tries I realised if I hadn't entered one, I could just bypass it by hitting return, right? RIGHT! It then took about 30 mins to update everything (with not a lot going on to let you know it's working ) but all was well. Except the fans now were ramped up to 30% at all times and those fans are proper turbines. The update of the BIOS/LCC/DRAC had cleared the previous fan auto scaling settings and now they had defaulted to failsafe - as the box now detected the Perc controller as a non-Dell LSI card it does this to ensure nothing melts down. The only real way to fix this is using IPMITOOL from a console session. It is NOT intuitive to say the least but in the end I got the fans to behave and the box is now essentially finished. BTW... Some of you will wonder why I didn't take the easier route and do this all from the DRAC. Well these old versions of the web server onboard don't have modern TLS protocols and simply will not load on modern browsers. If you have an older browser installed you might be able to do it but using the ISO flash drive method circumvented the whole problem. Anyway I now have a nice box that's not too loud, only takes 70w from the wall even when everything is busy and has lots of nice enterprise features like a DRAC, redundant power supplies and proper hard drive bays. It's cost me a grand total of £125 and some graft, but I think I'm set up for a few years at least

Well, I really want to know when Matrox cards are going to be supported... 😉

NetApp drives are formatted with 520 sectors, not the 512 that you need. There are Linux tools available that allow you to reformat them, look here:

Not a bad choice of HBA at all. Do you have active cooling on it? The 16 port cards get hot as hell when pushing a lot of data, the 8 port ones are bad enough. Could cause throttling if you don't have a fan on it. remember, these were designed for the hurricane force cooling inside servers.

AFAIK you can't passthrough an AMD GPU, only nVidia.

Can we get some clarification on this? I keep getting redball drives, always on disk 4 despite having changed the drive twice, the sata cale and the PSU in my server. I guess it's a RC5 issue, but having to redo parity (a 27 hour operation) every time it happens is becoming very tedious. The procedure listed in this thread doesn't work, I get the same issue as Athuruga, when I type: /root/mdcmd check NOCORRECT I get /root/mdcmd: line 11: echo: write error: Invalid argument and have to start a parity check all over again. Any chance of a 'Trust array' button sometime soon?