Re: Cache Replacement Algorithm

From: Eran Balter <baltere@dont-contact.us>
Date: Sat, 5 Aug 2000 13:54:10 +0200

Hi Henrik,

Thanks for the fast response!

When I asked about hybrid replacement policy, I ment the Hybrid replacement
policy, as proposed in the article "PROXY CACHING THAT ESTIMATES PAGE LOAD
DELAYS"
This policy takes into account 4 parameters:
- Document's size
- Number of references since last entered the cache.
- Latency of starting a connection to the document's origin server.
- Band width computed when downloading document from that server.

In order to implement this algorithm, I have to maintain a data-structure
for bandwidth's and latencie's to the origin servers. Moreover, I have to
add two fields: bandwidth and latency for each document entry. Before
getting into implementation, I would like to know if there is something
similar present at the original Squid source code, or somewhere else on the
planet :~`

Best regards,
Eran.
Received on Sat Aug 05 2000 - 05:00:07 MDT

This archive was generated by hypermail pre-2.1.9 : Tue Dec 09 2003 - 16:54:44 MST