Re: [squid-users] Hard Drive Latency

From: Matus UHLAR - fantomas <uhlar@dont-contact.us>
Date: Tue, 31 Aug 2004 09:59:01 +0200

On 30.08 17:43, James Gray wrote:
> Thanks for the info. Actually we have a 4 drive setup; 4x18Gb 10k RPM
> U160 in RAID 0+1 on a Compaq SmartArray5i controller with 64MB cache,
> Xeon 1.2GHz (IIRC) with 512MB ECC RAM :). Performance isn't such and
> issue for us, but redundancy is. Sure we could've gone RAID5 with one
> hot spare and probably got lower latency but we decided throughput was
> more important. Our bench-marking showed we got fractionally better
> throughput with RAID 0+1 over RAID 5 - the trade-off was slightly higher
> latency. Besides the whole thing is over-engineered by several orders
> of magnitude; it only supports 15 concurrent users on a 4Mbps link!!

If you know, what does RAID5 mean, you will never use it for squid cache.
RAID5 has slower writes in normal mode (it needs to read data and
calculate XOR's for writing!) and slower reads in degraded mode (it needs
read data and XOR's and calculate another data for reading) which happens
if one of disks fails.

If you want to use RAID with squid, you can join 2 and 2 disks to 2 RAID1
(mirrors) and use them as separate cache_dirs.

using raid0 (even in rais 0+1 or 1+0) for squid is useless and might even
slow things down.

-- 
Matus UHLAR - fantomas, uhlar@fantomas.sk ; http://www.fantomas.sk/
Warning: I wish NOT to receive e-mail advertising to this address.
Varovanie: na tuto adresu chcem NEDOSTAVAT akukolvek reklamnu postu.
BSE = Mad Cow Desease ... BSA = Mad Software Producents Desease
Received on Tue Aug 31 2004 - 01:59:08 MDT

This archive was generated by hypermail pre-2.1.9 : Wed Sep 01 2004 - 12:00:03 MDT