Re: redirecting netscape downloads to local ftp ?

At 19:04 14/06/97 +0200, Andreas Reeh wrote:
>Hi Squiders,
>
>sorry if this has been asked before, but I found nothing in the FAQ, nor
>in the searchable archive at www.cineca.it (is it no longer working? Most
>text I searched for has not been found)
>
>Ok, here me problem:
>
>I have many users who think it's a good idea to download Netscape and
>other very big files not from the original server but from one of the 50
>mirrors they find :-(((
>So these big objects are almost every time fetched from one of the many
>mirrors instead frome the cache where they still are, but under an other
>URL.

Something that was talked about before, which would be rather cool in squid
v2, is implementing MD5 checksums on large objects. When we get a request,
and go to fetch it, we can (hopefully) find out the size of the object in
the HTTP header. When it is over a certain size (set in squid.conf?), we
can go to an internal list of MD5 checksums and objects in a hash table. If
we get a match on both checksum and byte count, it is bound to be the same,
and so we fetch the object from the cache.

The table could be generated by a pool of "md5server"s, which would get
told when an object has arrived, and its location, and it would go off and
generate the checksum (they could be nice'd to some configurable level),
and send a message to squid when its done. In that time, of course, the
object could have expired, and squid will then ignore it; otherwise, squid
adds it to its table.

Is this worth contemplating in a future version?

Jonathan L.

Origin UK,323 Cambridge Science Park,Cambridge,England. Tel: +44(1223)423355
------[ Do not think that every sad-eyed woman has loved and lost... ]------
April 12th! Ra!Ra!----[ she may have got him. -Anon ]-----April 12th! Ra!Ra!
Help fight spam! http://spam.abuse.net/ These opinions are all my own fault
Received on Mon Jun 16 1997 - 06:55:14 MDT