ROCK install stalled

Thanks for responding. To be clear, I am running Linux Mint from a USB stick. I installed L Mint on the internal SSD but am unable to start it. What I want to do is reformat the entire SSD, but I can’t figure out how to do that from L Mint.

Yes. I’m aware. I just don’t understand why you are using Linux Mint when I suggested DietPi. You won’t be able to boot from the NVMe drive until you properly clear it. I hope you’ll consider following the process I suggested. Until you do, there’s not much more I can do to help. Sorry.

Rufus is not required to flash the DietPi image to a thumb drive. Etcher will work fine. You are not trying to install DietPi on the internal drive of the NUC. You just want to boot DietPi and use the commands I provided to clear the internal NVMe SSD.

I’m competing with suggestions from @David_Snyder a bit but…

There are a few things that happen during a Linux installation and each distro touches these things a little bit differently.

  1. The install drive will be re-partitioned. A partition is logical set of blocks (space) on a drive to build a filesystem. For your purpose you want 1 partition for data and one for swap space. You don’t need to know what swap does or why its there.

  2. The install will properly format the partitions for their use.

  3. The install will set-up the drive to be bootable. This “setup to be bootable” will be dependent on the bios, the machine, and a few other things.

Ubuntu Desktop does all 3 of these in a completely hands-off way. Just do a default install of Ubuntu and make sure the machine works properly. The fact you cannot boot Linux Mint is worrisome but I don’t know how to troubleshoot a non-booting Mint system but a non-booting Ubuntu system we can work through if you get that far.

Reason I suggest this… You cannot troubleshoot a ROCK install. Roon doesn’t provide any way to do this so if its not working we need to verify things another way. Additionally, yes, we could use the Mint bootable USB to set-up your internal drive but that’s a fair bit more difficult than just letting Ubuntu sort this out for you by performing an install.

I’ll also add… DietPi is an excellent distribution but its goal is to be lightweight. I don’t think lightweight is what you want at this point. You want a heavyweight, full featured, something focused on easy to use and darn supports every driver and hardware configuration imaginable and that’s something like Ubuntu. You have a system that will not boot from its internal drive. Let’s use the shortest path to “fix that” and make sure there isn’t some hardware or bios setting reason for this. Ubuntu will get you there.

In unix the “root” account has superuser privileges. The account can destroy a running system and for that reason it is no longer advised to “login as root” and some distributions actually prevent you from doing this. Instead it is recommended to switch user do, or sudo, (I may have made up what sudo stands for). When you run “sudo”, with no arguments in front of a command, you are running that command as root which is why I recommended running:
sudo fdisk -l

That commands runs fdisk with superuser privileges.

Thanks for your help and for your patience. I’ve tried and failed three times to flash the DietPi Native .img file to USB using Etcher. Frustration level is pretty high right now so I’m going to set it aside for a while and watch some football.

Still not a peep from Roon tech support on this.

This was a response to another post, not yours, who suggested that I install Ubuntu to the internal drive.

Unfortunately you are not likely to get any official Roon tech support over the weekend.

However you do already have two excellent members on board who know their stuff so you are in good hands I feel.
Hope you get this resolved and that frustration level can take a breather!

Yeah, of course the hard-working Roon folks deserve a weekend; I get that and I don’t mean to be a jerk - it’s just that frustration showing through. For the record, I did start this help ticket on Wed.

This should have been a 30 min exercise, and I went in to it expecting to be enjoying my new toy right now for Saturday night listening. When you buy used equipment there’s always a risk. I’ve had plenty of successes, but I have had a few fails over the years. At this point, the ROCK/NUC experiment is looking like a disappointing write-off to me and I don’t think I can recommend it based on this one experience.

At least I’ve learned a few things about Linux, which I have never tried before.

A post was split to a new topic: Linux reference books

Okay, here’s where I am, and thanks to all for your help. I plugged a 2.5 SSD into the internal bay and got Ubuntu to install and boot up (Linux Mint would not). I also installed Ubuntu on the NVMe SSD and got that to boot up too. Both are unstable. They crash after 20 - 30 min of browsing and normal use. The NVMe drive is far worse than the 2.5. After the crash, the drives sometimes do or don’t show up in the BIOS boot list (F10 during the BIOS startup). After I go to the BIOS setup (F2 on this NUC) and click the Legacy boot switch on and off I can get the drives to show up and boot. One thing that happens is that just before the Ubuntu startup menu appears, there is some frightening looking machine language that scrolls by, too quickly to read. Having never dealt with Linux or with a NUC, I don’t know if this is normal but it doesn’t look right to me. I should mention: I did flash a BIOS update early in the process. The ROCK install instructions still recommend this, but I later read reports that it was a bad idea. Too late now, but it did appear to conclude successfully.

Anyway, the question I’m now seeking advice on is this: Does this sound like a motherboard problem, a BIOS problem, or a bad drive? Or maybe bad RAM? Is it repairable or do I have a $400 doorstop?

Thanks in advance.

1 Like

There are two memory slots on that machine, are they both populated?

If you have two sticks of memory try running with 1 and then the other. That may tell you if you have a bad memory module.

This could simply be a kernel panic. Depends on exactly whats going on the system may dump the memory to the screen to help diagnose where the panic occurs. Memory errors can cause a kernel panic.

It’s difficult, from this description, to pinpoint exactly what the issue is. The fact you’ve tried two different drives makes that less likely the issue. So, memory is next to troubleshoot so let us know if you can get the thing stable while running on a single stick.

1 Like

It does look like a kernel panic, based on my experience with such things in DOS/Windows. The RAM is a single 16GB chip, so only one slot is populated. I noticed when I opened the case that the RAM chip is some generic off-brand. I will see if I have some compatible memory hanging around somewhere, but that isn’t likely. I assume it will run minimally with no RAM, so maybe I’ll try that. I have a busy week at work coming up so I don’t know when I’ll get back to this.

Thanks for the helpful suggestions.

I do appreciate your taking the time and effort to offer assistance. I have learned some things that you might find informative:

/dev/sda was not the drive that I was trying to reformat. It was the USB drive from which I booted Linux Mint. The NVMe SSD that needed cleaning was dev/nme0n1, as I originally deduced.

After a bit of study I discovered that I was able to run ‘fdisk’ and ‘wipefs’ from
Linux Mint by prefacing the commands with ‘sudo’. With that, DietPi wasn’t necessary for my objective.

Sometimes rudimentary things that are self-evident to the educated can be revelatory to the ignorant.

1 Like

I’m glad to hear that you were able to make progress. DietPi would have eliminated the need to bother with sudo. Sorry that it didn’t work out. I’ve done at least a half-dozen NUC builds with these SO-DIMM modules and never had a problem. I do recommend using a pair of matched modules instead of a single since the NUC platform supports dual-channel memory: https://www.amazon.com/gp/product/B019MRBKYG/ (includes two 8GB modules)

1 Like

You might try running a memory test from a bootable USB drive, google memtest86 or memtest86+.

The memtest86+ is an updated version of the original, I’ve used both over the years.

1 Like

Thanks for the suggestion. i ran memtest86+ from the Ubuntu startup menu (not from a bootable USB drive) and it totally redlined. I ordered new RAM from Crucial using their “compatible upgrades” tool to make sure I was getting the right stuff for the NUC8. The new RAM didn’t solve the problem and also redlined in memtest86+. My conclusion is that this is a bad DIMM slot or a motherboard problem. I don’t think it is worth getting it repaired, but I will talk to a repair shop to see what they think. At least I can harvest the 16GB RAM chip and 512GB NVMe drive (both Toshiba) in case I want to try again. And I’ve learned a few things about Linux from the experience.

2 Likes

I think it’s great that besides problems you always publish the found solution or the further solution steps here for the community. Thank you for letting us learn with you.

Any chance the NUC will boot with RAM in the other slot?

I tried all permutations of two chips in two slots and nothing worked. I suppose it’s unlikely that both slots are damaged, so it probably isn’t that.