Jump to content
Sign in to follow this  
stomp

HDD and unRAID crashed at every disk replacement

5 posts in this topic Last Reply

Recommended Posts

Posted (edited)

Hello,

 

So today I tried replacing one of my disk with one bigger. I noticed at reboot that something was wrong because the new disk made constant noise/bip and unRAID was logging errors on this particular disk. I already experienced several faulty disks with that specific Seagate models so I was not so surprised: I will end up buying WD from now on, lesson learned.

 

Now in the meantime, I tried to put my old little disk in the slot. unRAID got stuck mounting disk1 (not the one I'm trying to replace). I rebooted the server. unRAID got stuck at disk2, disk1 was showing "not installed" I think. Reboot after disable automount: now parity1 is "faulty) in maintenance mode.

 

To sum up, and this already happened last time with me losing data on 2 disks, everytime I try to replace one disk with one another, something goes wrong. And aside from the faulty new Seagate drive, unRAID do not behave normally as parity1 is now faulty (and I haven't even touched it).

 

You will find the diagnostic below.

 

What should I do? As I said, unRAID do not behave normally. It already happened once, now twice. Plus in the past, when adding disks, I also add problem with already installed disks not being recognized anymore (and needed a few reboots).

 

Can someone explain to me what is happening to drives that I don't even touch?

 

Thanks.

tower-diagnostics-20180413-1931.zip

 

EDIT: currently running xfs_repair -v on the faulty parity1 disk and it already showed "bad superblock" which looks bad because that is exactly what happened last time I lost 2 drives...

Edited by stomp

Share this post


Link to post

You rebooted before getting diagnostics so nothing in the syslog. Typically what happens when replacing a disk people will disturb the connections on another disk.

 

You say this happened before. Did you ask for help then? If it was very similar to this then you shouldn't have lost data.

 

The way forward is going to be set a New Config with the original disk back in. So put the original disk back in, check all connections, then post a new diagnostic.

Share this post


Link to post

Thanks for your quick reply.

 

I put the original disk back and started the array in maintenance mode. Here is the diagnostic. Tell me if I should do New config before diagnostic, it wasn't clear for me.

tower-diagnostics-20180413-1959.zip

Share this post


Link to post

Your diagnostics show only parity and disk8 disabled. Since you have dual parity you should be good to rebuild both of the disabled disks. The only thing that concerns me is disk2 has a lot of reallocated and should be replaced, and disk3 has quite a few also.

 

4 hours ago, stomp said:

EDIT: currently running xfs_repair -v on the faulty parity1 disk and it already showed "bad superblock" which looks bad because that is exactly what happened last time I lost 2 drives...

 

You can't repair the filesystem on parity because parity doesn't have a filesystem.

 

Stop the array, uncheck Maintenance mode, then post a screenshot of Array Operations.

 

I'm going to see if @johnnie.black has anything to add to this. Please don't do anything else without further advice.

 

Share this post


Link to post
Posted (edited)

Like @trurlmentioned some disks have reallocated sectors, and since they are the infamous ST3000DM001 should they should be replaced sooner rather than later, but they should be OK for the rebuild.

 

I assume the WD20EARZ is the original disk8, if so SMART looks mostly OK, there are several UDMA CRC errors, which unless they are old point to a bad SATA cable, but it would be safer to rebuild disk8 to a new disk in case something goes wrong during the rebuild, especially since you're down to one parity, so I would recommend replacing disk8 with a new one (edit: also replace the SATA cable to rule it out), rebuild it and at the same time resync parity1.

 

 

 

Edited by johnnie.black

Share this post


Link to post

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
Sign in to follow this