I came home last night to everything being offline thanks to not one but two drives failing. Not sure if it was heat, age, etc but I went in, stopped the array, shut down the server and removed the dead drives. After one very expensive trip to the store today, I'm now trying to replace the old dead drives (4TB) with two new ones (10TB). It was all they had so I grabbed them both. So now my parity (8TB) doesn't want me to start an array with a drive larger than it (I get that). I can't swap parity to one of the new 10TB drives, it just results in the "Too many wrong and/or missing disks!". I can't seem to find a configuration that this thing likes. What's the point of having the parity drive to save me in the event of a drive (or two) failure if I can't recover from it because too many drives are missing? Aside from rebuilding from scratch and losing more data than I care to think about, what are my options?

1 hour ago, DMills said:

What's the point of having the parity drive to save me in the event of a drive (or two) failure if I can't recover from it because too many drives are missing?

Single parity only allows you to recover one disk at a time. Parity contains none of your data. Parity (whatever system it is used in) is just an extra bit that allows a missing bit to be calculated from all the other bits. Parity plus all other disks are required to recover a missing disk. Dual parity would allow you to recover 2 missing disks at once.

 

1 hour ago, DMills said:

removed the dead drives

How do you know they were dead?

 

Plug the drives back in and attach Diagnostics to your NEXT post in this thread.

9 hours ago, trurl said:

Single parity only allows you to recover one disk at a time.

Now that I read that, and my head isn't buried inside the server case, it makes sense, so thanks for that

 

9 hours ago, trurl said:

How do you know they were dead?

Many reboots, many failed array starts, and then tried both drives in two other machines here. One would not even finish booting with either of the drives attached, the other with an external drive dock didn't detect either of them when plugged in. So I'm calling them dead.

 

A friend is bringing over a 4TB drive later today that he pulled out of his array due to read errors, we're hoping it will limp along enough to get mine to a state where I can at least swap the parity and then look at adding in the other 10TB drive.

 

I was getting warnings for months on one of the failed drives, read errors galore. The other reported errors maybe twice I think in the last month, maybe 3 times. I chose to ignore them, so it's 100% my fault and for months was entirely avoidable. You always think you have the time to get new drives before they become a problem, until you don't!

 

Thanks for you help and clarification!

 

D

55 minutes ago, DMills said:

A friend is bringing over a 4TB drive later today that he pulled out of his array due to read errors, we're hoping it will limp along enough to get mine to a state where I can at least swap the parity and then look at adding in the other 10TB drive.

I don't see how another bad drive can possibly help. It isn't merely the number of missing disks. Any other disks you try to put in isn't going to be the disks it is looking for when it decides you have too many missing disks. Parity is in sync with exactly those missing disks, and you can only rebuild one disk as long as all those other disks are present.

 

55 minutes ago, DMills said:

I was getting warnings for months on one of the failed drives, read errors galore. The other reported errors maybe twice I think in the last month, maybe 3 times. I chose to ignore them, so it's 100% my fault and for months was entirely avoidable.

Yes it was entirely avoidable. Are any of your other disks showing SMART warnings on the Dashboard page? Setup Notifications to alert you immediately by email or other agent as soon as a problem is detected. Don't let one problem become multiple problems and data loss.

 

I'm afraid you must consider the data on those disks lost if they truly can't be read.

 

Do you have backups of everything important and irreplaceable?

4 minutes ago, trurl said:

I don't see how another bad drive can possibly help.

Because in theory, the drive should still be usable, short term to get the array started. I realize it's not the same drive, but it's the right size. My biggest issue right now is the drives I want to use are bigger than my parity. This way at least I can try and get this working. Right now it's an expensive flower pot.

 

Yes I have backups, and I've written the data off as gone, I just have no place to restore the backups to yet. As for SMART warnings, yes, as I said, I was getting warnings for a while now, just thought I had more time than I did.

4 minutes ago, DMills said:

Because in theory, the drive should still be usable, short term to get the array started. I realize it's not the same drive, but it's the right size.

Bad theory.

 

It still won't let you start the array no matter how many other disks you try of whatever size. Your current disk configuration has missing disks. Those exact same disks (except one missing since you have single parity) must be present for you to use your current disk configuration.

 

Tools - New Config will let you set a new disk configuration. It will let you use whatever disks you do have exactly as they are and rebuild parity. Be sure you don't assign a data disk to a parity slot or it will be overwritten with parity.

1 minute ago, trurl said:

Tools - New Config

That will likely be what I have to do then.

Parity isn't magic. It isn't even very complicated. Understanding parity helps make more sense of how Unraid works.

 

https://wiki.unraid.net/Manual/Overview#Parity-Protected_Array

1 minute ago, DMills said:

That will likely be what I have to do then.

I don't see any other way forward. You have lost the data from the 2 disks you have been ignoring.

 

20 minutes ago, trurl said:

Are any of your other disks showing SMART warnings on the Dashboard page? Setup Notifications to alert you immediately by email or other agent as soon as a problem is detected. Don't let one problem become multiple problems and data loss.