Home / server / Questions / 449738
Accepted
cdecker
Asked: 2012-11-18 09:20:54 +0800 CST

Replace RAID1 Disc - things to consider

  • 772

One of my RAID controllers is complaining about a harddrive failure:

-- Controller informations --
-- ID | Model
c0 | PERC 5/i Adapter

-- Arrays informations --
-- ID | Type | Size | Status | InProgress
c0u0 | RAID1 | 465G | Degraded | None

-- Disks informations
-- ID | Model | Status
c0u0p0 | S13TJ1KQ503996 SAMSUNG HD502IJ 1AA01110 | Failed
c0u0p1 | S13TJ1KQ503997 SAMSUNG HD502IJ 1AA01110 | Online, Spun Up

so it looks like I have to replace the first disc in this array. But before ordering a new one I wanted to ask what I have to take into consideration.

Can I just order a new harddisc with 500+GB? What happens if I put a larger disc in? And finally will the RAID controller simply rebuild the mirror on the new disc or do I have to start it manually?

raid

2 Answers

  1. Ideally, you would replace the drive with the same drive (capacity, speed, etc). You could replace the drive with a large capacity drive or higher spin rate, however, the disk would only use the amount of space available on the lower capacity disk. Your RAID controller should automatically rebuild the disk.

    As mentioned in the comments, buy at least two and get it replaced ASAP!

    • 0
  2. Best Answer

    Can I just order a new harddisc with 500+GB?

    Yes. Make sure you order one which is at least equally in size or larger. This is not entirely trivial because not every 500 GB equals another 500 GB. These rounded values, not precise one.

    ( Your original disc shows as 465 GB ).

    What happens if I put a larger disc in?

    If you put in a larger disc (e.g. a 1GB drive) then the RAID controller will only use part of that drive. E.g. if I replace the first drive with a 1GB model. you would get this:

    Disk1:  465G GB in RAID | 500GB unused
disk2:  465GB in RAID

    And finally will the RAID controller simply rebuild the mirror on the new disc or do I have to start it manually?

    Most hardware RAID controllers will do this automatically. I expect your PERC5 to be one of them.

    As for @Michael Hampton comment. You can find some nice background information if you Google for bathtub curve and drive failure. Basically same models of a drive, used in the same way have a much larger chance of failing at the same time. If possible get drives from different batches to populate your RAID arrays.

    • 0