The picture says it all. One of the drives in a RAID-1 array failed and I had a replacement drive with the same model number ready.

When I tried to rebuild the array using the web interface it fails because the new drive is slightly smaller! Do I have any options here?


  • Generally speaking you can't rebuild raids with smaller disks, you'll need a disk of the same size or bigger. – FabioCH Apr 12 '13 at 19:19

I have done this in the past and just replaced it with a larger drive. Is there any reason you can't use a 250gb version of another drive in the same manufacturer or model?

That will work as I have done it many times.

Example I had some 1.5TB drives fail, I just replaced with similar model drives but 2TB versions. Rebuild worked just fine. Correct me though if I am wrong.

Yes this was all done with 3ware raid controllers under raid 1 and raid 6 using 9650se 3ware part number as the hardware raid controller.

  • Looks like this is the only option if I want to keep the server online. Thanks – Lucifer Sam Apr 12 '13 at 11:57

Backup files from the degraded volume. Delete the volume. Create a new volume. Restore the files.

Next time create volumes a few hundred MiBs smaller than disk size to avoid this situation.

  • This is a hardware RAID controller. Isn't the volume size irrelevant? – Lucifer Sam Apr 11 '13 at 23:51
  • 1
    I am talking about volumes (AKA logical disks) you create in the raid configuration tool. Obviously it is important for the disk to be big enough to fit all required data. And you can avoid this by making your volume slightly smaller than physical disk size – Alex P. Apr 11 '13 at 23:55
  • Thanks. I was thinking about OS volumes. I just ordered a larger hard drive for this issue but I'll definitely heed your advice when configuring new arrays in the future. – Lucifer Sam Apr 12 '13 at 11:59

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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