Hi all,

 

I have had this problem for awhile now, but I havent had the time to continually pursue it since the rest of my data is unaffected.

 

I had a TimeMachine Share set up called "MachineOfTime". Everything was fine until I started having problems and could not longer backup to it. I tried to delete the sparsebundle to no avail. Since the share is not empty, I can not delete the share. Luckily, this share is confined to a single disk (disk 5). My thought is that I could somehow remove all the data on the disk and prevent unRaid from rebuilding the disk.

 

How can I safely remove all of the data on disk 5? I dont need to, nor want to keep the data on the disk. I have already tried using the "rm -r" command on the .sparsebundle (through the user folder system) and it did not work (I let it run for 24+ hours, there was no disk activity, and all of the disks spun down including disk 5).

 

I am open to any method suggested. It seems as though I will have to take a more aggressive approach (could i use preclear.sh?)

rm -rf

 

It will take a long time.

rm -rf

 

It will take a long time.

Would it be quicker to dd zeroes to the md device and let unraid format it afterward?

There shouldn't be any need to zero the device at all, provided you already have valid parity. It would be quicker to just issue new filesystem on the device, where mdX is the device you want to format:

 

mkreiserfs /dev/mdX

 

 

Thanks BRiT! That is exactly what I was looking for!

 

Do I run "mkreiserfs /dev/mdX" with the Array Started or Stopped?

 

Are there any steps afterward to have unRaid recognize the disk?

Thanks BRiT! That is exactly what I was looking for!

 

Do I run "mkreiserfs /dev/mdX" with the Array Started or Stopped?

 

Are there any steps afterward to have unRaid recognize the disk?

 

Well, I tried it with the array stopped and it didnt work. Did I do something incorrectly?

 

root@Harvard:/mnt# mkreiserfs /dev/md5
mkreiserfs 3.6.21 (2009 www.namesys.com)

A pair of credits:
Vladimir Demidov wrote the parser for sys_reiser4(), the V3 alpha port, part of
the V3  journal  relocation code,  and helped  Hans keep  the business  side of
things running.

Elena Gryaznova performed testing and benchmarking.


Stat of the device '/dev/md5' failed.root@Harvard:/mnt#

Thanks BRiT! That is exactly what I was looking for!

 

Do I run "mkreiserfs /dev/mdX" with the Array Started or Stopped?

 

Are there any steps afterward to have unRaid recognize the disk?

 

Well, I tried it with the array stopped and it didnt work. Did I do something incorrectly?

 

root@Harvard:/mnt# mkreiserfs /dev/md5
mkreiserfs 3.6.21 (2009 www.namesys.com)

A pair of credits:
Vladimir Demidov wrote the parser for sys_reiser4(), the V3 alpha port, part of
the V3  journal  relocation code,  and helped  Hans keep  the business  side of
things running.

Elena Gryaznova performed testing and benchmarking.


Stat of the device '/dev/md5' failed.root@Harvard:/mnt#

the array must be started, otherwise the /dev/mdX devices will not exist.

I ran into the same problem and simply used mc to delete the remaining files.

 

Sent from a mobile device, sorry for any typos.

 

 

My standard cheat sheets for Midnight Commander. Issue the MC command from Telnet, or Console.

 

http://www.thegeekstuff.com/2008/10/midnight-commander-mc-guide-powerful-text-based-file-manager-for-unix/

http://www.trembath.co.za/mctutorial.html

http://linuxgazette.net/23/wkndmech_dec97/mc_article.html

 

I am still getting errors with mkreiserfs

 

'/dev/md5' looks mounted.Use -f to force over
root@Harvard:/mnt# mkreiserfs /dev/md5 -f
mkreiserfs 3.6.21 (2009 www.namesys.com)
...
'/dev/md5' looks mounted.Continue (y/n):y
Guessing about desired format.. Kernel 3.9.11p-unRAID is running.
reiserfs_create: could not open /dev/md5: Device or resource busy

 

Am I supposed to unmount disk 5? Sorry for being a noob

 

EDIT:

 

So I unmounted disk 5 (with the array still running) and ran mkreiserfs and it seems to be stuck on the following

Guessing about desired format.. Kernel 3.9.11p-unRAID is running.
Format 3.6 with standard journal
Count of blocks on the device: 488378624
Number of blocks consumed by mkreiserfs formatting process: 23116
Blocksize: 4096
Hash function used to sort names: "r5"
Journal Size 8193 blocks (first block 18)
Journal Max transaction length 1024
inode generation number: 0
UUID: 7b3f240d-dbef-41a7-90f8-139f73836699
Initializing journal - 0%....20%....40%....60%....80%....100%

 

This process is supposed to be pretty quick right?

 

EDIT 2:

 

It only took a minute or two

 

Syncing..ok
ReiserFS is successfully created on /dev/md5.

 

HOWEVER, the web gui is saying that there are unformatted drives present (which makes sense). Can I safely let unRaid format the disk? It does not explicitly say which drive it will be formatting which makes me wary.

It should mount properly if you do a clean shutdown and restart. Unraid has a nasty habit of calling all unmounted or unmountable volumes unformatted, so I'm not surprised it thinks it needs to format the drive since you unmounted it.

 

Come to think of it, I wonder if that might have been a shortcut to accomplish what you wanted. Unmount the md device, and let the gui mount and format it.

 

Might be interesting to do on a test server to see what happens.

Yes, I did a clean restart and everything seems to be working. Many thanks!