[SQU] Disk access as Squid bottleneck?

My Squid (v2.3S4) cache has plenty of CPU and RAM available to it. Also,
the hard disk on which the cache is stored has twice transfer rate of the
100Mbps network that the data crosses. Given these facts, am I correct in
thinking that disk access time is the bottleneck for my cache's throughput?

My hard disk is a very good IDE drive, but it has an IDE drive's access
time: 8.5ms. With sub-5.0ms SCSI drives on the market, I wonder if it
makes sense to get one as a replacement for my current drive. I'm willing
to spend the money, but I want to feel confident that the reduced access
time will really be noticeable as improved cache performance.

Any thoughts on this topic?

*** Steve Snyder ***

--
To unsubscribe, see http://www.squid-cache.org/mailing-lists.html