Home / server / Questions / 91776
In Process
Michael Kelley
Asked: 2009-12-08 10:12:08 +0800 CST

How to get at the extra drive space when upgrading a set of RAID 1 hard drives

  • 772

My current configuration is a Dell PowerEdge SC440 server with 2 SATA 160 GB drives in RAID 1 configuration. I bought two 1 TB drives and I want to replace the 160 GB with these 1 TB drives. There is only 1 partition on the virtual drive.

The first thing I tried was to remove 1 160 GB drive and replace it with a 1 TB drive. The 1 TB synched correctly, so I replaced the other 1 TB drive and that synched successfully. I was hoping I could extend the partition or at least create a new partition with the unused space. Unfortunately the Disk Management snapin only sees the virtual drive as 160 GB. I tried use Gparted, but it only recognized the virtual drive as 160 GB.

My question is what is the recommended way to acquire the extra space in the 1 TB? Can I take one of the 160 GB drives out, stick it in another computer and clone that drive, stick that back in the server, stick the other 160 GB drive in and sync that with the newly cloned driver? Or is there another easier way to do what I need to do?

I would prefer to solve this problem with freely available software.

hard-drive

4 Answers

  1. Does this server have a PERC or other RAID controller?

    Check the RAID controller configuration to see if you can grow the hardware RAID volume.

    Then use gparted to grow the NTFS into the new space.

    • 1

  2. The safest and easiest way to do do this would be to:

    1. BACKUP YOUR DATA!
    2. Pull both 160GB drives out (now you have 3 backups for all intents and purposes)
    3. Put the 1TB drives in
    4. Create the RAID-1 with the 1TB drives
    5. Restore drives to new Array
    • 1

  3. Last time I had to do this (PERC 5/i), I had to install the Dell software (OpenManage Server Administrator) in order to grow the volume. There was no option for it in the controller setup interface - I think I even called Dell to find out what I was missing and they told me there was no other way to get it done.

    • 1

  4. If you can activate a broken mirror, and your 160GB mirror set is still intact, you can:

    • Backup the existing data.

    • Shutdown. Insert both 1TB drives. Clear all RAID metadata from the 1TB drives (delete all volumes from them via RAID bios), and create a new 1TB volume.

    • Shut down, remove one 1TB drive, and insert one 160GB drive. If possible, block-read the the 160GB drive first to ensure all blocks are readable.

    • Activate both volumes in the RAID bios to yield two visible degraded mirrored volumes.

    • Boot gparted and use it to clone the 160 to the 1TB, and expand the filesystem on the destination if desired.

    • shut down and remove the 160GB. Insert the other 1TB. Reboot and allow RAID controller to rebuild the 1TB mirror (from the new clone).

    This method will avoid writing to your 160GB mirror source drives, and you will have one half of the mirror safely outside of the machine during the process... (assuming you are confident that half doesn't have undetected bad blocks...)

    The biggest risk will be inflexible RAID bios that might not allow two volumes at ones, or might arbitrarily "update" the metadata on one drive to match the other, etc...

    • 0