caching using refresh_pattern? force pages to be cached

Hi,

I'm trying to work through the problem of caching certain pages. I have

a proprietary web server which has very poor performance and at present

consider that caching may be a good way to resolve this. the server is a

web front end to a database, I am not concerned with keeping up to the
second accurate data.
As I understand it in order for squid to cache a page certain fields
must be present in the header.

HTTP/1.1 200 OK
Date: Tue, 23 May 2000 15:58:01 GMT
Server: Apache/1.3.9 (Unix)
Cache-Control: max-age=300
Expires: Tue, 23 May 2000 16:03:01 GMT
Last-Modified: Tue, 23 May 2000 15:22:42 GMT
ETag: "11405a-78d6-392aa242"
Accept-Ranges: bytes
Content-Length: 30934
Connection: close
Content-Type: text/html

The above header is one where the page is cached.

This is what is returned from the server that I wish to cache:

HTTP/1.1 200 OK
Content-type: text/html
Server: SWS/2.7.40
Connection: close
Date: Tue, 23 May 2000 17:09:20 GMT
Last-modified: Tue, 23 May 2000 17:09:20
GMT

I don't get Cache-Control, Expires or ETag, and presumably these are
required for normal cacheable pages. Working under 2.2 Stable 5.

My first question is am I making the correct assumptions? Second, has

anyone out there found a solution for this?

This is the config:
refresh_pattern -i http://winifred.dcu.ie 0 20% 360 override-expire
override-lastmod reload-into-ims ignore-reload
If there is no expire field in the header would there be a problem with

override-expire? Wasn't sure what the other options precisely do so
tried it without them and then with, no difference as far as I can see.

Also what does the 20% do?

Any ideas or hints would be gratefully received.

Karl.

Karl Devitt
Systems' Group
Computer Services Department
Dublin City University

Telephone: +353 1 704 5083
Email: karl.devitt@dcu.ie
Received on Tue May 23 2000 - 10:46:34 MDT