Re: how easy to recover RAID5?

mechphisto@xxxxxxxxx staggered into the Black Sun and said:
I'm looking to RAID5 together about three 600GB drives

Verbing weirds language! (-:

Core (7 at the moment, but I anticipate using 8). (Come to think of
it, we may end up using CentOS. Either way....)

The problem with Redhat-based things is that they're usually much more
of a pain to upgrade in place than Debian-based things. But if you're
cool with that, then it's easily doable to make a softRAID-5.

I'm having trouble finding information on how to recover a RAID 5.
I take it of one of the three drives goes out, the entire array is
unusable until the drive is replaced, right?

Um, no. All levels of RAID except 0 provide protection against a
single-disk failure. One disk fails, the array is still usable--
often slower, but usable.

But then what? You pop in a new drive, boot up--something
automagically happens to recover the data and redistribute bytes onto
the new drive? Or do I have to boot up with Knoppix/FC console disk
and run something?

No need to boot using a rescue system provided that /boot and / are
still usable. The last softRAID-5 I dealt with didn't recover
automagically. I had to fdisk /dev/sdc, create a 250G partition of type
0xfd on it, then "mdadm /dev/md0 --add /dev/sdc1" . It took 72 minutes
to resync, but was totally fine and usable before, during, and after the
resync.

There's a RAID HOWTO or at least an mdadm HOWTO (tutorial?) with a bunch
of examples of how to do various things using mdadm. HTH,

--
Quantum theory states that you have changed the actual contents of
this message by observing it. Since you have changed it, we can't be
held responsible for anything it may contain.
Matt G|There is no Darkness in Eternity/But only Light too dim for us to see
.

Relevant Pages

  • SVM RAID problem
    ... I have a RAID5 volume with 7 drives, ... running a metareplace on one of the maintenance devices and now it is ... Resync in progress: 0.0% done ... The resync has been stuck at 0% for a couple of hours and I've also ...
    (SunManagers)
  • Re: MD/RAID time out writing superblock
    ... We know this isn't caused by a single faulty piece of hardware, ... If it's multiple machines, it's much less likely to be faulty drives, ... The frequency of these timeouts ... up to 15-20 seconds during resync. ...
    (Linux-Kernel)
  • Re: limits on raid
    ... space as 'unused' and initialize it as it starts useing it?) ... it could zero all the drives first. ... I still fail to see the reason to actually "resync" the drives. ... array that has not been resync'd (or initialized as some hardware raids call ...
    (Linux-Kernel)
  • SV: SV: SV: [opensuse] Software RAID..
    ... I guess the main issue here is that my host is quite busy running ... My resync has now been running a little over an hour, ... These drives are not of the same brand, nor the same size, they're only ... I'll pull the disks and put them in another machine which is idle, ...
    (SuSE)