I was attempting to upgrade my array with a new drive.  I have a Yottamaster PS500C3 enclosure.

 

There are 4 drives in play. 1 X 14 TB Parity, 1 12TB Data and 1 14TB Data. There was an unreliable WD Drive (not in array, just in available pool)  in the mix as well that has been hooked up to the system, but not part of the arry. For the record, I never got the new drive installed, because when I put it in (and switched with the WD that I thought was bad, everything went to shit). So these drives above are what I originally had installed in the array. Including the bad WD - because I am uncertain whether the "bad" WD drive was what I pulled out or whether it was a drive originally in the array and the "spare" WD is in slot 1 or 2. I have not touched either of those and don't intend to.

 

I have 5 slots. Slots 1,2 have been untouched. The issue arose with slots 3,4,5. I had 2 drives in those two slots. I removed what I thought was the unreliable WD drive. When I placed the new drives in all hell broke loose.

 

Through much shuffleing, trying to get the drives back to their original order, I've managed to put myself in a situation where the raid is now reporting "    You may not add new disk(s) and also remove existing disk(s)."

 

I'm seeing 3 of 4 disks. I don't know what's what, because the Yottamaster won't tell me what drive is where. 

 

What is my best approach here? The raid will not allow me to start up because it's reporting "Invalid Expansion - You may not add new disk(s) and also remove existing disk(s)."

 

Is my best course of action to create a new config and assign all drives as data and see what comes up? I'm freaking out to say the least.

 

I'm seeing Disk4 as what it WANTS to assign as my parity, but I'm not sure how these Disk #s are derived when using this enclosure. Worried I will accidently toast a data drive because it was assigned as parity. I've moved the lower disks around and am just not confident where things are

 

I'm freaking out here, because drive assignments seem to be F^%$ 

 

Thanks in advance to anyone and everyone that can help. I am normally really careful with Unraid and I am just so ashamed I screwed this up so royally. And no - I don't have a backup of my flash drive..Such an IDIOT!

 

 

 

 

 

Edited April 23, 20251 yr by filmjock

1 hour ago, filmjock said:

I'm not sure how these Disk #s are derived when using this enclosure

Neither is Unraid. One of many reasons USB not recommended for assigned disks.

 

Attach Diagnostics ZIP to your NEXT post in this thread.

 

I'm definitely going to look into a non-USB solution after I get this sorted. Truly appreciate whatever advice you can all give. Diags attached

tower-diagnostics-20250422-2000.zip

Is the unreliable WD disk still working? If yes, connect it and post new diags.

It's included in the diags that I sent. I had plugged it back into the Yottamaster to try to get things back to where I started.

Goodness me ... first thing is don't panic!!

 

I am absolutely no expert with this, but the key thing is stay calm and work the problem. The information in the diagnostics is so limited has the USB enclosure has not passed any meaningful information, and now the drives appear to have been inadvertently shuffled.

 

If I was in your position, I would look to establish what data is on what drive. Depending on how you have formatted the drives, the beauty an Unraid array is it simply saves data on the drive and then uses another drive to check the data.

- Pull all your dives and put a sticker on them so you can make notes (Party, drive 1, other etc).

- Make a Linux bootable USB (Ubuntu for example), boot your machine using the bootable drive

- Put one drive in the caddy and plug into the machine

- Check in linux what is on the disk and make notes

- The parity should show nothing/unformatted

 

If you are lucky, you will end up with an unreadable parity disk, a 12tb and 14tb drive from the array (as you will have need the files) and then 3 other (the WD disk that started your problems and the other disks which were in slot 1 and 2).

 

Then if you really lucky, just put the parity, the 12tb and 14tb disks in and allocate them as disk 1 and 2 ... if it fails try allocating them the other way around.

 

If all else fails, maybe try and backup/consolidate your data onto a single drive and completely rebuild your array/Unraid getting a better way to connect your drives. For example get a cheap HBA card (I paid £30 on ebay for an LSI copy and so for it works well (for now)).

 

Yes, your enclosure is very good but I feel pushing all those drives over USB is not a good way to go.

 

Above all - best of luck!!

 

(ps - it looks like the Yottamaster just passes tray position regardless of what's in it, where a proper drive connection would pass your drive same and serial number which is why your disk assignment is FUBAR)

Edited April 23, 20251 yr by MrAndyBurns
PS

1 hour ago, filmjock said:

It's included in the diags that I sent.

There are only 4 disks connected, which one is missing?

The original configuration was 3 drives, 12Tb, 14Tb, and parity (14Tb). The problem WD was not part of the array and basically just listed as an unassigned disk device. The drive I was trying to add is not included in anything at this point. I just want to get back where I started and don't want to confuse myself.

 

Would Love additional thoughts on other potential approaches on this.

 

12 minutes ago, filmjock said:

The original configuration was 3 drives, 12Tb, 14Tb, and parity (14Tb).

If that configuration is still valid, I would recommend doing a new config with only those 3 disks to start with, do you know which one was parity, based on the model or serial?

I don't know which drive is my parity. I have 3 14s in play for parity and 2nd data (the 3rd 14 is the WD unreliable disk).

 

If I do a new config, wouldn't that essentially wipe my data drives? I hope I'm wrong...

 

Also - I'm seeing an x next to one of the disk assignments showing contents are emulated----how does that come into play?

 

And thanks again for all this. I owe all of you a debt of sincere gratitude!

 

 

Edited April 23, 20251 yr by filmjock

A new confit doesn't wipe any drives, the only danger is if you assign a previous data drive as parity, but if you had a disabled drive it could be a different story, I'm assuming these disks are OK and parity was valid with them:

 

40 minutes ago, filmjock said:

The original configuration was 3 drives, 12Tb, 14Tb, and parity (14Tb). The problem WD was not part of the array

 

If this is correct, and If you just don't know which drive is parity, attempt to mount all the drives with UD in read only mode, the one that doesn't mount is parity, then do a new config with them, you can check parity is already valid, again assuming it was, but should still then run a parity check.

I am concerned about the disk that says it's emulated. Currently in slot disk 2. If I create a new config, will I potentially lose that if it's a corrupted disk? Also - if I move it from the drop down list in the array and put it into UD with read only, will that mean that any emulation is now toast?

 

Also -

1)if I create a new config and nail the disk assignments, will parity need to rebuild or will it just "work"?

2)If the new config works, what happens to docker containers, etc? Will they disappear?

3)What about VMs that I've installed?

4)What happens to my cache drive? It has data on it.

 

Sorry for all the questions. Just trying to go into this with as much information as I can to make this as easy as possible.

Edited April 24, 20251 yr by filmjock

Did you use the array after the disk got emulated? In particular, were there any writes to the emulated disk?

19 hours ago, filmjock said:

I am concerned about the disk that says it's emulated. Currently in slot disk 2. If I create a new config, will I potentially lose that if it's a corrupted disk?

NO. Rebuild can't fix corruption anyway. But...

7 hours ago, JorgeB said:

Did you use the array after the disk got emulated? In particular, were there any writes to the emulated disk?

Any emulated writes would be lost.

 

--------------------------------------------------------------------

19 hours ago, filmjock said:

Also - if I move it from the drop down list in the array and put it into UD with read only, will that mean that any emulation is now toast?

NO. You don't even need to start the array to do this:

20 hours ago, JorgeB said:

attempt to mount all the drives with UD in read only mode, the one that doesn't mount is parity

-----------------------------------------------------------------

 

19 hours ago, filmjock said:

1)if I create a new config and nail the disk assignments, will parity need to rebuild or will it just "work"?

20 hours ago, JorgeB said:

do a new config with them, you can check parity is already valid, again assuming it was, but should still then run a parity check.

------------------------------------------------------------------

 

19 hours ago, filmjock said:

2)If the new config works, what happens to docker containers, etc? Will they disappear?

3)What about VMs that I've installed?

4)What happens to my cache drive? It has data on it.

New Config won't change any of the contents of any of your disks, unless:

20 hours ago, JorgeB said:

you assign a previous data drive as parity

 

 

12 hours ago, JorgeB said:

Did you use the array after the disk got emulated? In particular, were there any writes to the emulated disk?

No I didn't. It spooked me, nor did I attempt to spin up the array until I felt like I was on a good footing for fear of data loss.

Edited April 24, 20251 yr by filmjock

On 4/23/2025 at 1:56 PM, JorgeB said:

attempt to mount all the drives with UD in read only mode, the one that doesn't mount is parity, then do a new config with them, you can check parity is already valid, again assuming it was, but should still then run a parity check.

 

I'm considering getting a new system with an external facing SAS card and JBOD with individual SATA connectors. From everything I've read a setup similar to this is WAY more optimal than what I've been using.

 

The machine I'm using is a micro form factor dell with no PCI or external SATA connectors combined with a Yottamaster 5 bay enclosure. Am I playing with fire attempting to get this system back up and running when any time I attempt to expand the array could result in potential corruption of my drives? 

 

Want to know what you guys think is the smartest play here. When I started with Unraid, this setup seemed like a good start, but I'm rapidly realizing I need more processing power and reliability overall.

Edited April 25, 20251 yr by filmjock

Using SAS would be much better than USB, much less likely to run into issues.

One additional question: What happens to my cache disk?

If I do a new config, will it get toasted? if I drop it into a new machine will it get toasted?

Is there a smart way to back it up?

 

Thanks so much Mods and Friends!

21 minutes ago, filmjock said:

What happens to my cache disk?

On 4/23/2025 at 3:00 PM, filmjock said:

4)What happens to my cache drive?

On 4/24/2025 at 10:46 AM, trurl said:

New Config won't change any of the contents of any of your disks

 

 

I've got a new system inbound. Will keep you posted on progress. I truly appreciate all the advice so far!

Ok, so I've built the new system. 

 

New CPU, backplane, SAS card, etc

 

Setup a new config, but when I did, it blew out several of the settings in my docker containers. Emby, HomeAssistant and several others basically are acting like fresh installs.

 

The reason I started this whole process was because my data drives were full. I have found a few "good" settings folders on the data drives...drilling down individually, but I've had trouble restoring them in a way that works. I've tried creating new folders and copying the contents of the "good" folders to them using Krusader and the command line, but for some reason the containers come up with corrupted settings. Not sure if it's a permissions issue or what. Also of note, I attempted running the mover previously, but didn't have enough storage on the array to push the entire cache over, so I'm wondering if that also screwed me as well...

 

Would I be smart to use the move tool in the share browser window to move the "good" settings folders back to cache or a new folder entirely? Not sure what the best strategy is?

 

While working on the containers, I'm also running a parity check in the background - which is 60% complete and has corrected 1,033 sync errors so far on a 14TB parity. Once parity check is complete, I will add the new drive to the array and clear it. This will get me an additional 12Tb of space.

 

Any advice on how I can get my Dockers back up with "good" settings would be appreciated.

 

Thanks so much people! Getting There!!! You all really saved my bacon! Really can't thank you enough!

 

 

Edited May 6, 20251 yr by filmjock

post new diagnostics

Diagnostics attached

tower-diagnostics-20250506-1402.zip

2 hours ago, filmjock said:

I will add the new drive to the array and clear it.

If you haven't precleared the drive, Unraid will automatically clear the drive if you assign it to a new slot in the parity array, so parity will remain valid.

 

2 hours ago, filmjock said:

Would I be smart to use the move tool in the share browser window to move the "good" settings folders back to cache or a new folder entirely?

Disable Docker and VM Manager in Settings so you can work with these shares.

 

Then you should be able to use mover to move appdata and system from the array to cache. You need to change the direction of Mover action on appdata.

 

From your description, not clear what might have happened to some of your appdata. Assuming your docker templates specify /mnt/user/appdata, your containers should be able to find your appdata whether they are on cache or the array.