Re: Lockups and other trouble using large disks, raid(4), cgd(4), dk(4)

[Nino Dehne <ndehne%gmail.com@localhost>]

On Thu, Nov 20, 2008 at 10:13:40PM -0500, Thor Lancelot Simon wrote:
> On Fri, Nov 21, 2008 at 01:12:18AM +0100, Nino Dehne wrote:
> > 
> > OK, that went nowhere. I decided to try without cgd. I created an FFS
> > partition covering the whole raid0 using newfs -O 2 -b 65536 /dev/dk5
> > after changing the partition type of dk5 from cgd to ffs. The 64K
> > block size is necessary to match the 32 sectors stripe size of the RAID5.
> > Other values lead to abysmal write performance.
> 
> I do not doubt that under some condition you measured poor write performance.
> However, the explanation above is simply wrong.
> 
> Hint: how many kilobytes is 32 512-byte sectors?  But there's more wrong
> with the explanation than that.

I agree I phrased that poorly since it was just a side comment. The block
size should equal strip_size * number_of_data_components.

So for a 5-component RAID5 using a stripe size of 32 sectors the optimal
block size is 32 sectors * 4  = 64KB.

Also, file systems on a RAID5 should be aligned to that size, otherwise
write performance is equally poor.

Best regards.


-- 
Of course it runs NetBSD.