Re: [squid-users] Who is maintaining the FAQ?

From: Juri Haberland <list-squid.users@dont-contact.us>
Date: Tue, 11 Dec 2001 09:57:41 +0000 (UTC)

Henrik Nordstrom <hno@marasystems.com> wrote:
> We are here, only a little slow on responding.

I thought so - never mind.

> What was it you wanted to change in the FAQ?

Ok, here we go once again:
I stumbled across section 7.2 of the Squid FAQ and I believe it is wrong.
It states that in order to find the biggest object in the cache you have
to run
"sort -r -n +4 -5 access.log | awk '{print $5, $7}' | head -25".
IMHO this wrong, because
1) this doesn't take into account that some objects might not come from
the disk cache and also that some of those objects may never be cached
because
  a) they are too big or
  b) they match the 'no_cache' tag
2) this is only a snapshot of what actually is processed until you never
rotated your access.log since you started squid the first time.

egrep "(TCP|UDP)(|_(REFRESH|IMS|MEMORY))_HIT" access.log | sort -r -n +4 -5\
| awk '{print $5, $7}' | head -25
resolves 1) but 2) is still a problem.

The best solution would be to search the cache dirs, sort the results
according to size and somehow retrieve the URL from the files.

    Comments?

Cheers,
Juri

-- 
Juri Haberland  <juri@koschikode.com> 
Received on Tue Dec 11 2001 - 02:57:43 MST

This archive was generated by hypermail pre-2.1.9 : Tue Dec 09 2003 - 17:05:18 MST