Kernel 3.7.9 RAID6 hot replace success

Kernel 3.7.9 RAID6 hot replace success

[Brad Campbell]

G'day all,

I have a RAID6 of 10 2TB drives. One started to grow bad sectors at an 
alarming rate, so rather than fail and replace the drive I thought I'd 
use the hot replace method.

Add spare to array
echo want_replacement > /sys/block/md0/md/dev-sdm/state
Wait 6 hours
Remove failed drive
Job done.

Last time I tried this (lots of kernel versions ago) I had all sorts of 
issues with removing the failed drive and other sticky stuff that 
required rebooting to clean up. This "just worked".

Regards,
Brad

Re: Kernel 3.7.9 RAID6 hot replace success

[Roy Sigurd Karlsbakk]

> Add spare to array
> echo want_replacement > /sys/block/md0/md/dev-sdm/state
> Wait 6 hours
> Remove failed drive
> Job done.
> 
> Last time I tried this (lots of kernel versions ago) I had all sorts
> of
> issues with removing the failed drive and other sticky stuff that
> required rebooting to clean up. This "just worked".

This is great. Any idea when that came in?

Btw, with as many as 10 drives in a RAID-6, I'd keep at least one spare permanently there…

Vennlige hilsener / Best regards

roy
--
Roy Sigurd Karlsbakk
(+47) 98013356
roy@karlsbakk.net
http://blogg.karlsbakk.net/
GPG Public key: http://karlsbakk.net/roysigurdkarlsbakk.pubkey.txt
--
I all pedagogikk er det essensielt at pensum presenteres intelligibelt. Det er et elementært imperativ for alle pedagoger å unngå eksessiv anvendelse av idiomer med xenotyp etymologi. I de fleste tilfeller eksisterer adekvate og relevante synonymer på norsk.
--
To unsubscribe from this list: send the line "unsubscribe linux-raid" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: Kernel 3.7.9 RAID6 hot replace success

[Brad Campbell]

On 06/03/2013, at 2:34 AM, Roy Sigurd Karlsbakk <roy@karlsbakk.net> wrote:

>> Add spare to array
>> echo want_replacement > /sys/block/md0/md/dev-sdm/state
>> Wait 6 hours
>> Remove failed drive
>> Job done.
>> 
>> Last time I tried this (lots of kernel versions ago) I had all sorts
>> of
>> issues with removing the failed drive and other sticky stuff that
>> required rebooting to clean up. This "just worked".
> 
> This is great. Any idea when that came in?
> 
> Btw, with as many as 10 drives in a RAID-6, I'd keep at least one spare permanently there…
> 

Im pretty sure I last tested it about 3.2 or  3.3.

I do keep a permanent spare drive, but with 3 arrays I manually allocate it in the rare event it is required. This drive had not failed out of the array, so the spare sat idle until I proactively replaced the dodgy drive.--
To unsubscribe from this list: send the line "unsubscribe linux-raid" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: Kernel 3.7.9 RAID6 hot replace success

[Keith Keller]

On 2013-03-05, Brad Campbell <brad@fnarfbargle.com> wrote:
>
> I do keep a permanent spare drive, but with 3 arrays I manually allocate it in the rare event it is required. This drive had not failed out of the array, so the spare sat idle until I proactively replaced the dodgy drive.--

Are you using the spare-group attribute of arrays in mdadm.conf?  If so,
how well has it worked in your scenario?

--keith

-- 
kkeller@wombat.san-francisco.ca.us

Re: Kernel 3.7.9 RAID6 hot replace success

[Brad Campbell]

On 06/03/13 10:21, Keith Keller wrote:
> On 2013-03-05, Brad Campbell <brad@fnarfbargle.com> wrote:
>>
>> I do keep a permanent spare drive, but with 3 arrays I manually allocate it in the rare event it is required. This drive had not failed out of the array, so the spare sat idle until I proactively replaced the dodgy drive.--
>
> Are you using the spare-group attribute of arrays in mdadm.conf?

No, have not got that advanced yet. I just keep on top of the monitoring 
e-mails and manage things manually. Thanks for the reminder though. I do 
need to have a look at that a bit closer when I get some more time.

Would not have helped in this application though.

Regards,
Brad