So what you're saying is I can't even try to repair the filesystem right now? Is there any possibility that the drives are responsible for the disconnect? I've already reseated everything and about the only thing left to do is replace cables and/or controller. That will take days and this thread will be lost in the meantime.

You're not any worse off than you were when you did New Config to get that disk1 removed.

 

You can try to repair the filesystem on physical disk4 if you can get it to connect and stay connected.

19 minutes ago, Goreliann said:

this thread will be lost in the meantime.

When you make a new post to the thread it rises to the top and will show up as unread to me.

Well I've just shutdown, reseated Disk 4, and rebooted. Disk 4 still says contents emulated, but after attempting a filesystem check, it couldn't complete it and gave an error, but a few minutes later it says Disk 4 returned to normal utilization and all my files came back.

I'm now tempted to swap the drive out and try to rebuild off the parity and just accept whatever is lost.

On 1/28/2024 at 12:01 PM, trurl said:

00:1f.2 SATA controller [0106]: Intel Corporation 82801IR/IO/IH (ICH9R/DO/DH) 6 port SATA Controller [AHCI mode] [8086:2922] (rev 02)
	Subsystem: Super Micro Computer Inc 82801IR/IO/IH (ICH9R/DO/DH) 6 port SATA Controller [AHCI mode] [15d9:d780]
	Kernel driver in use: ahci
	Kernel modules: ahci

01:00.0 RAID bus controller [0104]: HighPoint Technologies, Inc. Device [1103:2720] (rev 03)
	Subsystem: HighPoint Technologies, Inc. Device [1103:0000]
	Kernel driver in use: mvsas
	Kernel modules: mvsas

 

My guess would be that the Intel controller is on the motherboard, and the HighPoint controller is a separate card.

 

You should be able to figure out where the cables go to determine which of these controllers disk1 and disk4 are connected to.

All drives save the jump drive are connected through the card. It has two ports and two breakout cables connected to it totaling 8 drives.

1 minute ago, Goreliann said:

I'm now tempted to swap the drive out and try to rebuild off the parity and just accept whatever is lost.

This is absolutely not an option. No reason at all to think parity is valid.

 

And really no reason to think both disk1 and disk4 are bad. SMART for both of these disks looks good. This is almost certainly a problem elsewhere, and you can't make any progress until you fix it.

 

1 minute ago, Goreliann said:

All drives save the jump drive are connected through the card. It has two ports and two breakout cables connected to it totaling 8 drives.

How many ports on the motherboard? Can you at least get disk4 connected that way?

1 minute ago, trurl said:

How many ports on the motherboard? Can you at least get disk4 connected that way?

It's been so long since I put this thing together, I'm not sure. I suppose I could. Right now I'm thinking just replace the card and cables. Then what? Try to rebuild Disk 4 onto itself again? Then if that works, new config and rebuild parity? Is that what I'm looking at?

If parity were valid, there would be no reason to rebuild it, and if it isn't valid, you can't rebuild disk4 from it.

 

I suppose you could try to rebuild disk4, and hope against hope that parity by some miracle was valid. But I wouldn't attempt that onto the same disk.

 

The current contents of disk4 are really all you have that can be trusted, except for those other disks that haven't been involved.

 

We need to see if physical disk4 needs to have its filesystem repaired, and if so, repair it. And accept that as the best we can do unless you have backups. Then we can rebuild parity.

Right now I'm at the point that although the contents are emulated, the files on Disk 4 are present. That would suggest to me that the parity, for most intents and purposes anyway, is functioning even if the emulated contents are incomplete. I can live with that. So it seems to me that my best step now is to replace the card and cables, physically replace Disk 4 with another drive and try to rebuild it. Does that sound reasonable?

How are these drive cages powered? Have you done anything to make sure that is working?

They're powered off a single standard PSU. None of the other drives have had any issues though so I would assume that if power was an issue, that all drives in a cage would fail. Additionally, Disk 1 and Disk 4 are in different cages, but those seem to be the only drives that are disconnecting.

If power is borderline for some reason, some disks might have more difficulty than others for various reasons, especially if all are running. And everything we have been trying to do requires all disks to run so requires more power than simply reading or writing files.

So maybe while I'm replacing thing, I might as well replace the PSU too. I imagine the 16TB drives use more power than the 4TB, but I figured there's still enough power coming out of the thing. I did reseat the power connect6ions when I reseated all the drive cables. The problem for me is all this hardware is pretty old so it may be cascade failing. Oy vey. I'll start replacing stuff on a shotgun approach and hope that ONE of the things I replace, if not all, will allow me to rebuild a full array someday.

14 minutes ago, Goreliann said:

I imagine the 16TB drives use more power than the 4TB

Age more than capacity might be the determining factor. Newer tech is often better in multiple ways.

59 minutes ago, Goreliann said:

the contents are emulated, the files on Disk 4 are present. That would suggest to me that the parity, for most intents and purposes anyway, is functioning even if the emulated contents are incomplete

Good point.

 

When you get things working again, you can rebuild to another disk, such as that disk you had assigned as disk1, and keep original disk4. And repair filesystem if either need it. Then you will have 2 versions of disk4 you can work with.

OK. So now I've swapped out the raid card and the cables. No when I fire up unraid, it says all my devices are missing. It's not recognizing any drive. Every slot says no device and doesn't give me any option to select one. Sigh.

The card I swapped is an HBA so maybe the old one wasn't. I honestly don't remember. I saw something on the forum about a new config, but that has me a bit nervous.

The post on the forum also said something about rebuilding the disks one at a time. Will that be possible when one was emulated to begin with?

For further info, I had a friend of mine suggest I reboot the array and look for a config tool for the new card. Maybe that's an option?

It seems to me, what I need to do is Tools>New Config....Check parity is valid, start the array. 

 

Is that right? I think that will initiate a parity check, no? I just want to make sure this is the right thing to do before I do it.

Well since I didn't hear from anyone. I hit New Config. Now on the main screen everything is listed as unassigned and no disks are showing up at all. It's well and truly borked, isn't it?

I think I've managed to configure the new raid card. Once I rebooted after doing that, unraid can now see all the drives. I assigned them to the exact same Disk locations as before the new config and started the array. It's now doing a parity check. Let's hope to God that everything is ok after that.