Take the 2-minute tour ×
Server Fault is a question and answer site for professional system and network administrators. It's 100% free, no registration required.

I have a centos server that runs a whm/cpanel with mysql, apache, bind and email daemons. The server now has 95% of the server disk full. It is a dell poweredge 1950 with 2 spaces for SAS disks. I have it setup now with mirrored raid with the two disks currently in the server ( SAS 146GB 10K drives ).

I need to get more space available on the server but I'm unsure about how to go about that. What would you guys suggest for getting more space? I have purchased two 600GB SAS 10k drives to replace the 146GB drives ( i want to run the new larger drives in mirrored configuration as well ).

I've thought about trying to pull one of the disks in the mirrored configuration and putting a larger disk in and letting the system try to rebuild the array. However I'm worried that this could cause problems and that the larger disk may not even be accepted into the array. If the raid array became corrupted or unusable, then I'd be stuck doing a bare metal restore from backups which I'd like to avoid like the plague.

I've also thought to just put these larger disks in another free poweredge 1950 I have laying around and then move the data from the existing server over to this server. I have been experimenting with this offline, however I'm not sure this is the best way to go about it? The method I was going to go about this was

  • boot both servers off a live cds
  • on the destination server with larger disks, i would format the new 600GB disk with partitions and a file system, then mount the root partition. I would setup rsync daemon to be able to push to the root partition
  • from the source server with smaller disks I would mount the root drive, then rsync -avh everything up to the destination server new root partition
  • from the destination server I would then need to modify fstab, grub boot configuration if necessary, rebuild the initrd if necessary then do a grub-install to push grub onto the new disks

If all went well with the above I'd have a duplicate server with larger drives, and if that didn't work I'd still have the old server with smaller disks to revert back to in emergency.

I've experimented with some computers laying around my house, I got as far as having the new destination boot up but because of different hardware the destination server kernel panics, appears to be a missing kernel module for the hard drive controller. I was going to look into resolving this and seeing if my little experiment works so I can replicate on the live server, however I was wondering if I am going about this the wrong way?

Any suggestions are appreciated?

Thanks

share|improve this question
    
Is the RAID software or hardware? –  MadHatter May 18 '12 at 15:46
    
@MadHatter hardware raid, it's a dell perc 5/i, thanks –  Mike May 18 '12 at 15:51
    
And don`t forget to use LVM in the future. Then you will have more flexibility with regards to resizing "partitions". –  Nils May 18 '12 at 20:53
    
Mike, then I'm afraid I'm with Basil. Dump, replace, restore. –  MadHatter May 19 '12 at 5:14
add comment

migrated from stackoverflow.com May 18 '12 at 15:38

This question came from our site for professional and enthusiast programmers.

2 Answers

up vote 2 down vote accepted

The safest way would be to back your server up, turn it off, replace the drives, and restore to them.

share|improve this answer
    
would you mine elaborating? back them up = just do a hard copy of the contents of root partition or ?. Restore = repartition new disks, format file system, copy contents of root partition back and install grub? –  Mike May 18 '12 at 16:11
1  
I didn't get too specific because I don't know how you currently back it up. Personally, I'd do a bare-metal-restore backup of the existing drives to tape, swap the hardware, and then do a bare metal restore. I use netbackup, but I don't know what kind of backup software you have. –  Basil May 18 '12 at 17:34
add comment

Play with your spare 1950:

  • Insert both drives
  • use both drives a single targets
  • On first drive use your "old" partition scheme
  • Install OS on one drive (don`t forget space for the Dell Utility Partition)
  • Make "bigger" partitions on the second drive
  • Copy your data over to the second drive with rsync -axHS for each mountpoint
  • Remove first drive (or offline it with OMSA), swap drives, boot from rescue-cd
  • Install grub
  • Reinstall or online the "small" drive
  • Remove that drives raid-controller-configuration (OMSA or PERC5i-BIOS)
  • Try to build a RAID1 with the big and the free drive via OMSA

If this works - try to break up the virtual drive via OMSA into single drives again.

Hint: Put "Tikanga" into /etc/redhat-release" and you will be able to install and run OMSA 6.5.0 without problems.

  1. Break up your RAID1
  2. Insert bigger disk
  3. Rebuild
  4. Reboot with bigger disk
  5. Insert other bigger disk
  6. Rebuild to RAID1

I can imagine that this could work - but doing a restore to the new disks is probably faster.

share|improve this answer
add comment

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.