Jump to content

[SOLVED] been running on parity - second disk got disabled and now shows....


Recommended Posts

Hello,

 

The last 2 days i have been running on parity due to some capacity issues on the power supply. While i wait for the new power supply i have been running the server.

 

Today though, i noticed another disk get disabled. I rebooted the server and it shows as "unformatted".

 

I read the FAQ and waited about 20 minutes, still no go. shows unformatted.

 

How can I repair my disk4 without messing up my parity for the other drive?

 

In the meantime i'm running a reiserfs --check. I couldn't do /dev/md4 cause it recognized it as bad blocks so i'm running it on /dev/sdg1 which is disk4 (the one showing as unformatted)

 

Syslog:

 

Aug 16 22:12:32 kenny emhttp: disk4 mount error: 32
Aug 16 22:12:32 kenny emhttp: shcmd (24): rmdir /mnt/disk4
Aug 16 22:12:32 kenny kernel: REISERFS warning (device md4): sh-2006 read_super_block: bread failed (dev md4, block 2, size 4096)
Aug 16 22:12:32 kenny kernel: REISERFS warning (device md4): sh-2006 read_super_block: bread failed (dev md4, block 16, size 4096)
Aug 16 22:12:32 kenny kernel: REISERFS warning (device md4): sh-2021 reiserfs_fill_super: can not find reiserfs on md4
Aug 16 22:12:33 kenny kernel: REISERFS (device md2): found reiserfs format "3.6" with standard journal
Aug 16 22:12:33 kenny kernel: REISERFS (device md2): using ordered data mode
Aug 16 22:12:33 kenny kernel: REISERFS (device hdb1): found reiserfs format "3.6" with standard journal
Aug 16 22:12:33 kenny kernel: REISERFS (device hdb1): using ordered data mode
Aug 16 22:12:33 kenny kernel: REISERFS (device hdb1): journal params: device hdb1, size 8192, journal first block 18, max trans len 1024, max batch 900, max commit age 30, max trans age 30

Link to comment

Hello,

 

The last 2 days i have been running on parity due to some capacity issues on the power supply. While i wait for the new power supply i have been running the server.

 

Today though, i noticed another disk get disabled. I rebooted the server and it shows as "unformatted".

 

I read the FAQ and waited about 20 minutes, still no go. shows unformatted.

 

How can I repair my disk4 without messing up my parity for the other drive?

 

In the meantime i'm running a reiserfs --check. I couldn't do /dev/md4 cause it recognized it as bad blocks so i'm running it on /dev/sdg1 which is disk4 (the one showing as unformatted)

 

Syslog:

 

Aug 16 22:12:32 kenny emhttp: disk4 mount error: 32
Aug 16 22:12:32 kenny emhttp: shcmd (24): rmdir /mnt/disk4
Aug 16 22:12:32 kenny kernel: REISERFS warning (device md4): sh-2006 read_super_block: bread failed (dev md4, block 2, size 4096)
Aug 16 22:12:32 kenny kernel: REISERFS warning (device md4): sh-2006 read_super_block: bread failed (dev md4, block 16, size 4096)
Aug 16 22:12:32 kenny kernel: REISERFS warning (device md4): sh-2021 reiserfs_fill_super: can not find reiserfs on md4
Aug 16 22:12:33 kenny kernel: REISERFS (device md2): found reiserfs format "3.6" with standard journal
Aug 16 22:12:33 kenny kernel: REISERFS (device md2): using ordered data mode
Aug 16 22:12:33 kenny kernel: REISERFS (device hdb1): found reiserfs format "3.6" with standard journal
Aug 16 22:12:33 kenny kernel: REISERFS (device hdb1): using ordered data mode
Aug 16 22:12:33 kenny kernel: REISERFS (device hdb1): journal params: device hdb1, size 8192, journal first block 18, max trans len 1024, max batch 900, max commit age 30, max trans age 30

 

My advice - power down your server and wait for your new PSU.  Until you have a known powerful enough PSU you will never know if the symptoms you are seeing are PSU related or something else.  Whatever you do, do not press the format button!

Link to comment

Hello,

 

The last 2 days i have been running on parity due to some capacity issues on the power supply. While i wait for the new power supply i have been running the server.

 

Today though, i noticed another disk get disabled. I rebooted the server and it shows as "unformatted".

 

I read the FAQ and waited about 20 minutes, still no go. shows unformatted.

 

How can I repair my disk4 without messing up my parity for the other drive?

It is HUGE mistake to try any repair until you get the proper power supply.

 

In the meantime i'm running a reiserfs --check. I couldn't do /dev/md4 cause it recognized it as bad blocks so i'm running it on /dev/sdg1 which is disk4 (the one showing as unformatted)

 

Syslog:

 

Aug 16 22:12:32 kenny emhttp: disk4 mount error: 32
Aug 16 22:12:32 kenny emhttp: shcmd (24): rmdir /mnt/disk4
Aug 16 22:12:32 kenny kernel: REISERFS warning (device md4): sh-2006 read_super_block: bread failed (dev md4, block 2, size 4096)
Aug 16 22:12:32 kenny kernel: REISERFS warning (device md4): sh-2006 read_super_block: bread failed (dev md4, block 16, size 4096)
Aug 16 22:12:32 kenny kernel: REISERFS warning (device md4): sh-2021 reiserfs_fill_super: can not find reiserfs on md4
Aug 16 22:12:33 kenny kernel: REISERFS (device md2): found reiserfs format "3.6" with standard journal
Aug 16 22:12:33 kenny kernel: REISERFS (device md2): using ordered data mode
Aug 16 22:12:33 kenny kernel: REISERFS (device hdb1): found reiserfs format "3.6" with standard journal
Aug 16 22:12:33 kenny kernel: REISERFS (device hdb1): using ordered data mode
Aug 16 22:12:33 kenny kernel: REISERFS (device hdb1): journal params: device hdb1, size 8192, journal first block 18, max trans len 1024, max batch 900, max commit age 30, max trans age 30

The disk blocks cannot be read since the disks have (probably) shut themselves down from poor voltage regulation.  You really can't do anything until you get a more capable power supply.
Link to comment

Hello,

 

The last 2 days i have been running on parity due to some capacity issues on the power supply. While i wait for the new power supply i have been running the server.

 

Today though, i noticed another disk get disabled. I rebooted the server and it shows as "unformatted".

 

I read the FAQ and waited about 20 minutes, still no go. shows unformatted.

 

How can I repair my disk4 without messing up my parity for the other drive?

It is HUGE mistake to try any repair until you get the proper power supply.

 

In the meantime i'm running a reiserfs --check. I couldn't do /dev/md4 cause it recognized it as bad blocks so i'm running it on /dev/sdg1 which is disk4 (the one showing as unformatted)

 

Syslog:

 

Aug 16 22:12:32 kenny emhttp: disk4 mount error: 32
Aug 16 22:12:32 kenny emhttp: shcmd (24): rmdir /mnt/disk4
Aug 16 22:12:32 kenny kernel: REISERFS warning (device md4): sh-2006 read_super_block: bread failed (dev md4, block 2, size 4096)
Aug 16 22:12:32 kenny kernel: REISERFS warning (device md4): sh-2006 read_super_block: bread failed (dev md4, block 16, size 4096)
Aug 16 22:12:32 kenny kernel: REISERFS warning (device md4): sh-2021 reiserfs_fill_super: can not find reiserfs on md4
Aug 16 22:12:33 kenny kernel: REISERFS (device md2): found reiserfs format "3.6" with standard journal
Aug 16 22:12:33 kenny kernel: REISERFS (device md2): using ordered data mode
Aug 16 22:12:33 kenny kernel: REISERFS (device hdb1): found reiserfs format "3.6" with standard journal
Aug 16 22:12:33 kenny kernel: REISERFS (device hdb1): using ordered data mode
Aug 16 22:12:33 kenny kernel: REISERFS (device hdb1): journal params: device hdb1, size 8192, journal first block 18, max trans len 1024, max batch 900, max commit age 30, max trans age 30

The disk blocks cannot be read since the disks have (probably) shut themselves down from poor voltage regulation.  You really can't do anything until you get a more capable power supply.

 

Even if I unplugged the newly installed hard drives that caused the problem in the first place?

 

The reiserfs came back with no errors. I can wait a few more hours, the PS will be here today.

 

I'll keep you guys posted if i run into other problems

Link to comment

Using /dev/sdg1 can easily destroy any chance of using the parity to correct a failed disk. Messing with a bad power supply and having 2 disks show as disabled may have already gone too far to recover.

 

FYI, once a disk is disabled it won't just "come back" on it's own. Keep your fingers crossed both disks are OK because you'll likely have to invalidate the parity and rebuild the parity to get the server working correctly again.

 

Peter

 

Link to comment

Ok, I got the new PS plugged it in.

 

Still shows unformatted so I did the following;

 

 

mkdir /mnt/test

mount /dev/sdi1 /mnt/test (<-- This is unformatted drive).

 

Checked it out and all the files are still there PHEW! Now i'll do a smart check.

 

If S.M.A.R.T. passes I assume I can "Trust my Array"? http://www.lime-technology.com/wiki/index.php?title=Make_unRAID_Trust_the_Parity_Drive,_Avoid_Rebuilding_Parity_Unnecessarily

 

 

Link to comment

Not likely, and certainly not at all if you have written to the disabled disk.

 

I have no idea what you're doing to show the disk is not formatted. It won't just be put back into the array without you doing it. It's still not part of the array if you just installed the power supply and booted the machine.

 

Peter

Link to comment

Not likely, and certainly not at all if you have written to the disabled disk.

 

I have no idea what you're doing to show the disk is not formatted. It won't just be put back into the array without you doing it. It's still not part of the array if you just installed the power supply and booted the machine.

 

Peter

 

My bad, it shows disabled not unformatted sorry about the confusion. I don't mind losing the small amount of data i might have written over the last 24 hours as long as the rest of the data is there i'm ok with that.

Link to comment

It doesn't work that way. Just stop the array, unassign the disk, start the array, stop the array, reassign the disk and start again. The data will be rebuilt. Of course, this assumes that all the other disks are OK and that the disabled disk is still being emulated correctly.

 

Peter

 

Link to comment

It doesn't work that way. Just stop the array, unassign the disk, start the array, stop the array, reassign the disk and start again. The data will be rebuilt. Of course, this assumes that all the other disks are OK and that the disabled disk is still being emulated correctly.

 

Peter

 

 

That won't work though because it sees 2 disks with errors. It won't be able to rebuild data off 2 disks now.

 

I checked them both though and they are find when i mount them.

 

The problem is though that now one disk shows as "new" and the other as "disabled". (red ball and blue ball)

 

 

Link to comment

OK, stop the array, assign the drives as you want them and then log-in and type initconfig. Answer Yes. Start the array and let the parity be built. If the drives are good, then they'll all come back almost immediately. Don't do anything else like pressing the format button if it appears.

 

Peter

 

Link to comment

OK, stop the array, assign the drives as you want them and then log-in and type initconfig. Answer Yes. Start the array and let the parity be built. If the drives are good, then they'll all come back almost immediately. Don't do anything else like pressing the format button if it appears.

 

Peter

 

 

That did it. Thanks :) I was so nervous...

Link to comment

Archived

This topic is now archived and is closed to further replies.

×
×
  • Create New...