Re: Ideas/solutions to caching identical objects ...

From: Luka Frelih <luka@dont-contact.us>
Date: Mon, 2 Dec 96 05:15:13 +0100

> Is it possible (ie in a later version of squid, or the next ? :) to
> recognise a previously gathered copy and redirect references to the
> copy present in the cache?

it would be, but then you would need to parse all the files using
some hash function, remember the result and compare it with the
results in cache. this could lead to lots of burned cpu time.

but, on the other hand, the recognizing process could run
independently from the caching one, just comparing objects on disk.
this would lead to temporarily duplicated data, which would get
linked together a while after it arrives on disk.

i don't know what this would do to the code that manages updating
and expiry of objects, and especially to the on-disk database
routines.

L
Received on Sun Dec 01 1996 - 20:22:20 MST

This archive was generated by hypermail pre-2.1.9 : Tue Dec 09 2003 - 16:33:48 MST