Re: [opensuse] mdadm keeps breaking my array

On Friday 23 January 2009 18:57:48 Boyd Stephen Smith Jr. wrote:
On Friday 2009 January 23 12:14:42 Bob Williams wrote:


Does this mean that sdd is failing and needs to be replaced? I sort of
know the answer I'm going to get, but thought I'd check here first, in
case I was missing something obvious.

Could be cable, controller, disk, or even some rare kernel bug, but it
is probably the disk. If you can get the RAID1 sync'd and healthy
again, you might try, shutting down, swapping the cables, and seeing if
the problem follows the disk or the cable.

Actually, what I forgot to mention, is the two drives are attached to a
PCI SATA-RAID controller card. I rebooted the computer after posting my
last message, and on entering the card's setup, it appears the card has
also got these drives setup as a RAID1 array (I must have set this up
myself, sometime. Probably in a previous computer). I destroyed this
array, so the two drives are independent at this level, and rebooted.

This time, mdadm --manage is rebuilding the array again. After 90 minutes,
it's showing

#cat /proc/mdstat
Personalities : [raid1]
md0 : active raid1 sdd1[1] sdc1[0]
976759864 blocks super 1.0 [2/1] [U_]
[===>.................] recovery = 17.9% (174884480/976759864)
finish=550.0min speed=24294K/sec
bitmap: 124/466 pages [496KB], 1024KB chunk

unused devices: <none>

so I'll be optimistic, and assume the two RAID setups were tripping over
each other.

If it fails again, I'll try your suggestion.

Thanks.

Bob
--
Registered Linux User #463880 FSFE Member #1300
GPG-FP: A6C1 457C 6DBA B13E 5524 F703 D12A FB79 926B 994E
openSUSE 11.1, Kernel 2.6.27.7-9-default, KDE 3.5.10
Intel Core2 Quad Q9400 2.66GHz, 4GB DDR RAM, nVidia GeForce 9200GS
--
To unsubscribe, e-mail: opensuse+unsubscribe@xxxxxxxxxxxx
For additional commands, e-mail: opensuse+help@xxxxxxxxxxxx

Relevant Pages

  • Re: Question about RAID1
    ... hda: OS RAID1, swap, dataA ... No advantage in throughput, this is the simplest data redundancy configuration (one half of the mirror set can fail, the other set takes over as primary, replacement drives are automatically synchronised - which may hit performance). ... On your primary qwuestion of overall system performance, my suggestion is to partition your volume setup so that you can take advantage of the different setups. ... RAID6 uses dual-parity distributed across the array, hence offers extremely high fault tolerance as well as increased performance over a single drive or a RAID5 array. ...
    (comp.os.linux.misc)
  • Re: Debian installer
    ... Oh, and while I'm at it, how about raid1 arrays with more than 2 drives? ... 640 gig mirror and reported the array used all 4 drives. ... What I really wanted was to create a RAID10 array during the install and was playing around trying to see if I could work out another solution during the install even if it would be a little slower. ...
    (Debian-User)
  • Re: [opensuse] mdadm keeps breaking my array
    ... PCI SATA-RAID controller card. ... last message, and on entering the card's setup, it appears the card has ... array, so the two drives are independent at this level, and rebooted. ...
    (SuSE)
  • raid1, rh9 & SATA drives
    ... The drives for the raid setup are Western Digital hot-swappable Serial ... So I figured with hot swappable drives running raid1 I'd ... ping the machine, console was unresponsive, ctl-alt-del didn't work. ...
    (RedHat)
  • Raid & SATA drives
    ... The drives for the raid setup are Western Digital hot-swappable Serial ... So I figured with hot swappable drives running raid1 I'd ... ping the machine, console was unresponsive, ctl-alt-del didn't work. ...
    (comp.os.linux.hardware)