Intel RAID

Issues related to hardware problems
Post Reply
wesfeik
Posts: 1
Joined: 2020/02/16 04:16:30

Intel RAID

Post by wesfeik » 2020/02/16 04:17:59

Hello all,

I have a CentOS 6 server with onboard Intel RAID 5 with 5 Seagate server grade 500GB hard drives. The first drive (drive0 of the array) has failed (SMART message on power up)and as-is the server will no longer boot up. I am assuming that the boot sector used in this configuration is located on the first drive of the array, thus causing this behavior. All other 4 drives report that they are in good condition.

I received an exact model this week for a replacement, I was wondering if swapping out the drives and booting to the CentOS Live disk would work or be the best option to rebuild the array with the new disk. I have found a bit of info using Google, but none that seems to apply to my specific situation, and I really don't want to lose the current array by making a bad decision. Anyone out there had this experience? Any advice is appreciated.

User avatar
TrevorH
Site Admin
Posts: 33202
Joined: 2009/09/24 10:40:56
Location: Brighton, UK

Re: Intel RAID

Post by TrevorH » 2020/02/16 11:35:14

Welcome to the hell that is FakeRAID.

If this was my server and the data was important, I would get a very large external drive and image all of the remaining disks to it using dd before I started doing anything else at all. That would take a bit for bit copy of everything on each drive in case of problems later on but it will take a while.

Theoretically, in RAID 5, you should be able to withstand the loss of any one disk and still be able to boot up and use the system. I have no idea why you cannot. That boot sector should be spread across the other drives in exactly the same way as all the rest of the data on there. It should not have taken your boot ability away at all.

Otherwise I think you replace the failed disk then use the machine's BIOS to add the new drive into the array (I don't think it will do that automatically though it might). Since it's FakeRAID, unfortunately it will not rebuild on its own until there is an operating system present that knows about FakeRAID so, yes, you have to boot something on there to get it to rebuild. The CentOS install media has a rescue environment that might be used though it's command line only.
The future appears to be RHEL or Debian. I think I'm going Debian.
Info for USB installs on http://wiki.centos.org/HowTos/InstallFromUSBkey
CentOS 5 and 6 are deadest, do not use them.
Use the FAQ Luke

Post Reply