[clug] Is a weekly RAID scrub too much?
eyal at eyal.emu.id.au
Sun Feb 26 06:32:57 UTC 2017
On 26/02/17 17:15, Paul Wayper wrote:
> On 24/02/17 22:35, Eyal Lebedinsky wrote:
>> I was idly browsing when I came upon an announcement from Toshiba of a new NAS
>> HDD family:
>> I took a look at the spec
>> and there I see
>> "Rated Annual Workload (Total TB Transferred per Year, R/W) 180 TB/year".
>> and this is the same for the three models 4/6/8TB.
>> Doing the math: a weekly scrub will read 52*8=416TB a year, more than twice
>> the designed load.
>> Even a monthly scrub will use up more than half the workload.
>> Is the scrub doing more harm than good by shortening the service life of the
> So, what do you mean by "weekly RAID scrub"?
> I've never heard of anyone doing this on a RAID array.
> I've used 'scrub' on a disk or files to overwrite them with random data to
> avoid anyone using a scanning electron microscope on them to view your data
> again. With disk geometries, cylinder tolerances, and encoding mechanisms
> these days, even overwriting with zeros will render the data inaccessible to
> almost all attackers.
I use 'shred' for this.
> But I've never heard of 'RAID scrubbing'. What are you doing, Eyal?
RAID scrub refers to, in linux software RAID, a 'check' request. It means that
the full RAID is read and verified. For RAID6 it means P and Q are verified.
You start such a check with something like:
# echo 'check' >/sys/block/md127/md/sync_action
and then you can observe the progress with:
$ cat /proc/mdstat
Personalities : [raid6] [raid5] [raid4]
md127 : active raid6 sdi1 sde1 sdf1 sdh1 sdc1 sdg1 sdd1
19534425600 blocks super 1.2 level 6, 512k chunk, algorithm 2 [7/7] [UUUUUUU]
[======>..............] check = 31.7% (1241121264/3906885120) finish=564.2min speed=78742K/sec
bitmap: 1/30 pages [4KB], 65536KB chunk
> Have fun,
Eyal Lebedinsky (eyal at eyal.emu.id.au)
More information about the linux