On Sat, 12 Sep 1998, Stephen Baxter wrote:
> I was thinking of using ICP to make sure. This will greatly reduce false
> hits to the point where the most common reason they would happen is from a
> squid mesh having non common refresh patterns.
If digests are well implemented and tuned, the most common reason for false
hits should be no different. (The probability of a false hit due to the
collisions in the digest itself is small and can be further reduced if
needed).
> On digests (we haven't tried them yet) does anybody have any feedback on
> the CPU power and time to generate a digest
Theoretically, you probably do not care much about the time it takes to build
a digest as long as you finish before the next refresh period. If CPU is a
problem, the build process can be extended in time so it has negligible
impact on Squid performance.
> and what is the best time interval between digest updates ?
The conservative default is one hour. Unfortunately, we do not have enough
data to speculate about the optimal setting. The good thing is that your
cache controls how often to rebuild, and your peers would obey your policy.
Alex.
Received on Fri Sep 11 1998 - 17:17:21 MDT
This archive was generated by hypermail pre-2.1.9 : Tue Dec 09 2003 - 16:41:57 MST