-----Original Message-----
From: [EMAIL PROTECTED] <[EMAIL PROTECTED]>
To: Stephen Costaras <[EMAIL PROTECTED]>
Cc: [EMAIL PROTECTED] <[EMAIL PROTECTED]>
Date: Saturday, November 07, 1998 11:50 PM
Subject: Re: Improving RAID5 performance (what disks/configuration?)
>For random I/O, you want each access assigned to just one disk, so
>you need the per-disk stripe size to be at least as big as the reads;
>I *think* this is what the chunk size is, but I'm not certain. On the
>other hand, to do random writes fast you want the chunk size to be
>only slightly larger than the write size, or exactly the write size if
>you can...and if I understand RAID correctly, parity must be generated
>from n-1 of the n disks in the array and written to the other disk
>(which stores the parity for that stripe, rather than a portion of
>the data)...which means n-2 random reads for each small random write...
True, I have been calculating chunk size based on the mean I/O size *
the number of disks. Which falls into the 4-8k range I/O, across 8 disks
is where I got the 64k chunk size value.
>If you need fast random writes, perhaps RAID 0 would be better? If
>you need fast random writes and redundancy, RAID 0+1 might be the
>ticket, especially if you need random write speed and redundancy
>but don't need so much capacity.
Unfortunatly I need both size and speed. Ideally I'd like to use
RAID 3 but I can't find support for that unless I buy the hardware.
>IMHO you shouldn't waste money on intermediate speed drives. If you
>truly need speedy random I/O, you need the lowest seek time and the
>lowest rotational latency you can afford, especially as track switch
>times and track-to-track seek times rise from 0.5 ms (on the likes of
>the Ultrastar 2XP, Atlas II, or the early Barrcudas) to 3 ms on more
>recent 7200 RPM Barracuda models. Blame rising areal density (in the
>form of falling distances between tracks, requiring much finer head
>positioning), but that's the case. If all you need is transfer rate
>and capacity, the Viking II, Ultrastar 2ES, or latest low-end 7200
>RPM Seagate drive will give that to you.
Currently I have several (10+ Seagate Barracudas (older ST15150W) models
that I was trying to salvage for this use. These now are probably
intermediate
speed drives by your terms, I couldn't chuck them for cheetah's though
without some hard figures for backing.
>Remember with Bonnie to use top or some other tool to measure CPU usage,
>as time spent by raid5d (e.g.) isn't seen by Bonnie...but the key issue
>is certainly to characterize your need properly, then apply the best RAID
>arrangement to meet that need.
Right, I'm not using Bonnie to show CPU utilization I have other means
for that (ps -xau). Bonnie is just an easy way to quantatativly measure
random I/O. The application on this machine (INN w/ CNFS) is getting really
bogged down due to disk I/O. I need to find a way to optimize the system
as much as possible (for highest random disk I/O operations) with hopefully
retaining much of the previous investment in disks.
I'm trying to see what other people are able to push their subsystems
to for random I/O w/ RAID5 and how they have it laid out. I might just
be at the maximum for this configuration.
Steve
