Array data disk failed


Recommended Posts

Hi

I have an array with two 3TB disk, one data and one parity, and cache - 128GB SSD. Disk are encrypted (data + cache). Data disk is now marked as disabled and I want to replace it. My plan was to increase array size to 4TB per disk in first step. As I understand documentation (Storage Management | Unraid Docs https://docs.unraid.net/unraid-os/manual/storage-management/#parity-swap) and some threads on forum I should first change/swap parity disk. So please confirm:

1. I should stop array, shutdwon unraid

2. Insert new disk

3. Start unraid, add new, bigger disk to array, format it

4. Copy parity to new disk (there should be some new operation in Main -> Array Operation -> Copy)

5. After parity copying is done shutdown unraid

6. Remove old parity disk, remove old data disk,  insert new data disk

7. Start unraid, include new disk in array

8. Rebuild array

 

I should remove data disc - disabled now in 6 step or after step 8? Does encryption affects somehow whole process (excluding extra time for encpryption), bigger risk of failure while replacing or need an extra step? As I understand documentation encryption should be transparent so no extra risk. Please confirm.

 

--

  kajot

Link to comment

Many seem to think "format" just means "get a disk ready for use" (whatever that vague idea means).

 

"Format" actually has a very specific meaning, which is very important when using Unraid.

 

Format is a write operation that means "write an empty filesystem" to the disk. Many people have lost data from not understanding this simple meaning. When they are replacing a disk, they think then need to format it before rebuilding. If you format a disk in the array, Unraid treats this write operation just as it does any other, by updating parity. So, after formatting a disk in the array, the only thing that can be rebuilt is an empty filesystem.

 

Parity swap is the correct procedure to replace a disk while also upsizing parity. Would be better if you followed the instructions exactly.

 

If you want to do the preclear test on either disk, you should do it before anything else, with the disks as unassigned devices. Format has no place in any of this, it is either pointless, or dangerous.

 

Since disk1 is already disabled (and the emulated disk1 is mounted), you can skip the first 4 steps in the documented procedure.

Link to comment

Yesterday I was able to start parity copy operation and it was done Today.. After parity was copied I shutdown Unraid. Should I expect any trouble to resume the proccess after power up? In the parity swap tutorial there were no power off + power up between parity copy and array rebuild (but I assume this is not a problem at all).

 

I assume I can use other drive than my old parity one to rebuild array (I've already ordered one more new 4TB). Can you confirm?

 

While copying parity this was marked as unlocked (old parity disk which was source) - I didn't enter passphrase. I assume that this was because parity is just raw parity from other devices so it is kind of "encrypted parity" (I mean parity from encrypted disc content) and unraid do not keep somwhere the passphrase to encrypt array. 

Link to comment

So I need to again add only parity disks (new as parity, old as data), copy parity to data, and then rebuild array but then can I change the setup - I mean use second new disk instead of old parity one as new data drive?. Or I need to divide it for different steps:

1. copy and rebuild (new disk as parity, old parity as temporary data)

2. remove temporary data insert other brand new as data and rebuild again?

Link to comment

I'm not sure when and how something goes wrong but after parity copy when I was starting array the message about "Missing encryption key" was displayed. There were no input for passphrase.

I found a solution to put passphrase in /root/keyfile and after that start array again. Now rebuilding proccess is ongoing.

Is it way how it should be done - I mean in that case passphrase need to be specified like above?

Link to comment

ok, I think it is side effect of Parity swap actualy :) As I mentioned I plan to change two discs but it is not possible at once. I had only tree slots for disc so I temporary removed cache and it could be that my docker containers were there (do not remember now). I'm rebuilding data second time (I change finaly both dics new parity and new data and it was not possible to do it in one copy-rebuild operation, finally in my case I need copy parity, rebuild array /new parity + old parity disc as tmp data/ and second rebuild /old parity disc as tmp data changed to new data/). When I finish I try to summarize the proccess for people who may have similar issue.

Link to comment
1 hour ago, kajot said:

temporary removed cache and it could be that my docker containers were there

That is where they should be if you are doing it right.

 

So nothing to do with the parity swap procedure really.

 

You should disable Docker and VM Manager in Settings until you get cache back in. And post diagnostics before you enable them again. They probably got created on the array and you will need to clean that up first.

Link to comment

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Restore formatting

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.