Re: [squid-users] R: [squid-users] cache size and structure

From: Matus UHLAR - fantomas <uhlar_at_fantomas.sk>
Date: Thu, 2 Jul 2009 17:36:22 +0200

>> No, it is not. It may be 1mbps .OR. 76GB for week. it can't be anything
>> "per second for week". You may mean 1mbps during weekdays, 1mbps during
>> the whole week, 1mbps average.

On 30.06.09 19:06, Riccardo Castellani wrote:
> I want to say if weekly average of http traffic is 1 Mbps (monitored by
> mrtg tools), all http traffic, which goes to my squid, is 76 GB in a
> week.

Aha. So, you should have 76-152 GiB of disk cache on a few fast disks, if
you can afford that... along with a few gigs fo ram.

don't put too much of cache on one disk, it can slow you down. You can start
with e.g. 15GiB and increase by time, until you'll notice increase fo
response times. In such case, keep the L1 on 128.

> infact mrtg gives me these information:
> maximum peak for day
> traffic average for day
> ...for week
> ...for month
>
> Do you understand my calculates ?
>
>> ok, are your child caches configured as neighbours to each other?
>
> both squid (B,C) have configured as parent cache squid A
> I don't know what means "configured as neighbours to each other". Do you
> reference B as neighbours to C ?!

Exactly, and vice versa. If they can connect to each other. And, if they are
not behind slow links, although it's probably they are (otherwise there
would be no need for 3 caches, right?)

-- 
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.
Emacs is a complicated operating system without good text editor.
Received on Thu Jul 02 2009 - 15:36:27 MDT

This archive was generated by hypermail 2.2.0 : Mon Jul 06 2009 - 12:00:02 MDT