[Starcluster] enlarge volume

Austin Godber godber at uberhip.com
Thu Aug 12 17:52:19 EDT 2010


I am not sure where you ought to go from here.  There is clearly 
something wrong with the filesystem on that partition.  Without knowing 
the exact commands, its hard to know what when wrong.

Do you have a snapshot?  I have used parted and e2resizefs to expand an 
ext3 filesystem without trouble.


Actually, you can peek at the filesystem a little bit with the following 
two commands, but if I had to guess, I would say they are going to fail 
for the same reason:

blkid -s TYPE -o value /dev/sdz1
sudo tune2fs -l /dev/sdz1

Make sure you are providing the path to the partition and not just the disk.

Austin

PS - Apologies to moderators and others, I sent this mail earlier from 
another address.  So, sorry if you get it twice.


On 08/12/2010 05:29 PM, Dan Yamins wrote:
> Same error.
>
> resize2fs 1.39 (29-May-2006)
> resize2fs: Bad magic number in super-block while trying to open /dev/sdm1
> Couldn't find valid filesystem superblock.
>
> Something is corrupted on the drive when re-writing the partition table
>
>
>
> On Thu, Aug 12, 2010 at 12:18 PM, Justin Riley <jtriley at mit.edu 
> <mailto:jtriley at mit.edu>> wrote:
>
>     -----BEGIN PGP SIGNED MESSAGE-----
>     Hash: SHA1
>
>     On 08/12/2010 11:43 AM, Dan Yamins wrote:
>     > Does this completely confirm that the larger 500 GB volume is
>     completely
>     > "seen"?   (And if so, does this mean that nothing like resize2fs or
>     > xfs_growfs is necessary?)
>
>     Dan,
>
>     - From the fdisk -l output you sent you can see that you have a larger
>     volume now (536.8GB) however your partition is still only 20GB big:
>
>     20964824(blocks)/1024.0(blocks/MB) ~= 20GB
>
>     You need to recreate the partition to be 536.8GB and then run
>     resize2fs:
>
>     If you're doing this on a StarCluster you'll first need to run the
>     following command (as root) on each node of the cluster for the
>     mount_path defined in your volume's StarCluster config:
>
>     $ umount MOUNT_PATH
>
>     Then shutdown the nfs service (as root) on the master node:
>
>     $ /etc/init.d/nfs stop
>
>     After that do the following (as root) to resize your new volume:
>
>     $ umount /dev/sdz
>     $ fdisk /dev/sdz
>
>     # Type 'd' to delete the primary partition
>     # Type 'n' for new partition
>     # Type 'p' for primary
>     # Type '1' for 1st
>     # Type Enter for 1st cylinder
>     # Type Enter for last cylinder (full disk)
>     # Type 'w' to finish
>
>     $ e2fsck -f /dev/sdz1
>     $ resize2fs -p /dev/sdz1
>     $ fsck -f -y /dev/sdz1
>
>     Then you should be able to mount the new volume:
>
>     $ mount /dev/sdz1 MOUNT_PATH
>
>     And start nfs again on the master:
>
>     $ /etc/init.d/nfs start
>
>     And re-mount the NFS drives for that volume on each of the nodes:
>
>     $ mount MOUNT_PATH
>
>     HTH,
>
>     ~Justin
>     -----BEGIN PGP SIGNATURE-----
>     Version: GnuPG v2.0.15 (GNU/Linux)
>     Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/
>
>     iEYEARECAAYFAkxkHucACgkQ4llAkMfDcrnxIwCgjFOPtLte3TT59m8mPONqPf3O
>     ZhAAnRQ8ByIehbl6CS/8QbraCistzu6+
>     =9VLQ
>     -----END PGP SIGNATURE-----
>
>
>
> _______________________________________________
> Starcluster mailing list
> Starcluster at mit.edu
> http://mailman.mit.edu/mailman/listinfo/starcluster
>    


-- 
Austin Godber
godber at jumpbox.com
http://www.jumpbox.com

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mailman.mit.edu/pipermail/starcluster/attachments/20100812/a8d6e207/attachment.htm


More information about the StarCluster mailing list