Re: RAID 5 vs RAID 10 benchmark

From: Howard J. Rogers <hjr_at_dizwell.com>
Date: Fri, 26 Nov 2004 06:17:45 +1100
Message-ID: <41a62fcf$0$12900$afc38c87@news.optusnet.com.au>

Jan wrote:
> Several scientific (or logical) conclussions were already made about
> RAID 5
> that this is not the best option for write intensive databases. It
> seems very logical also to me when I read arguments against RAID 5.
>
> But I`m confused by following benchmark showing that it is not
> worthful anymore (at least for a tested storage) to thinks about RAID
> 5 vs RAID 10 from the performance point of view.
>
> http://www-1.ibm.com/servers/storage/disk/ess/pdf/raid5-raid10.pdf
>
> Jan

The concern with RAID 5 is not whether there is a write penalty so much. With these days of massive battery-backed cache, that is almost a non-event (though the "I" in RAID has, as a result, long since ceased to mean 'inexpensive').

The real concern with RAID 5 is what happens when you pull one of the disks out. A single physical I/O now requires a read to be made against every surviving member of the array, so that the missing data can be deduced. RAID5's I/O performance under failure conditions is woeful. And you suffer it for the duration of the array rebuild when you plug the disk back in.

Try the same thing with RAID 1+0... the loss of one disk makes no difference, since a complete array is still available. And plugging the disk back in means there's just one disk to re-silver from.

Regards
HJR Received on Thu Nov 25 2004 - 13:17:45 CST