RE: [squid-users] caching websites automatically

From: Amos Jeffries <squid3_at_treenet.co.nz>
Date: Thu, 12 Feb 2009 11:06:19 +1300 (NZDT)

>>Leonardo Rodrigues Magalhães schrieb:
>>>
>>> Amos Jeffries escreveu:
>>>> The big question is WHY BOTHER?
>>>>
>>>> Squid will cache whatever it can as it passes through, things get
>>>> cached as users request them. This saves you on bandwidth and disk
>>>> space. While giving followup clients the speed they would like.
>>>
>>> i completly agree with you. I really dont see the point on this
>>> ....... but just gave Lorenor a simple and feasible way for acchieving
>>> what he wants ........ despite the fact i also think that's stupid and
>>> undeeded to do this kind of procedure :)
>>>
>>> I know it sounds weird, but I need it for my bachelor thesis. Thanks
>>> for
>>> the help!
>
> Also, wouldn't it be of some use caching high-use high-media sites in
> off-peak hours (00:00-04:30 in my case) so any new images are already
> cached for the morning? On the link I use we have almost no activity
> between 6PM and 8AM, but between 8AM-6PM its heavily in use, when I
> started reading this thread I thought pre-caching sites at around 4AM,
> like the BBC, Sky News etc. would be of some use. Yes, I know these sites
> get updated regularly during the day, but at least some of the new content
> would be already in the cache.
>
> Am I missing something?

The expiry times, high-change or fast-reacting sites like those often have
low page times and may have expired between 4am and 6am so a set pre-fetch
time is not as good as it first seems.
If you analyse each popular site, you can often identify the time to
spider for best results. But it will differ between what sites your
clients prefer and where you are on the Internet.

Amos
Received on Wed Feb 11 2009 - 22:06:29 MST

This archive was generated by hypermail 2.2.0 : Thu Feb 12 2009 - 12:00:02 MST