Re: [squid-users] 2.4.20-aa and LARGE Squid process -> SIGSEGV

From: Ralf Hildebrandt <Ralf.Hildebrandt@dont-contact.us>
Date: Sun, 22 Dec 2002 10:06:21 +0100

* Joe Cooper <joe@swelltech.com>:

> Probably not. The performance difference between using 64 MB on a
> heavily loaded cache (20Mbps throughput, for example) and 128 MB is
> relatively small--perhaps 5%. The difference between 128 MB and 256 MB
> is even smaller. Doubling again is probably not even measurable.

I see. So we're investigating a non-problem!
 
> You probably aren't hurting yourself by configuring less than 700 MB. I
> never configure more than 256 MB, except in an accelerator with no disk
> cache. Your OS is going to cache disk accesses anyway, so the memory
> doesn't go unused.

Thanks for the clarification. So it boild down to "Does squid cache
better than the OS's kernel?"

-- 
Ralf Hildebrandt (Im Auftrag des Referat V a)   Ralf.Hildebrandt@charite.de
Charite Campus Mitte                            Tel.  +49 (0)30-450 570-155
Referat V a - Kommunikationsnetze -             Fax.  +49 (0)30-450 570-916
"Smith & Wesson - the original point and click interface." 
Received on Sun Dec 22 2002 - 02:06:25 MST

This archive was generated by hypermail pre-2.1.9 : Tue Dec 09 2003 - 17:12:11 MST