On Wed, Jun 26, 2002 at 12:13:15AM -0400, Scott Young wrote:
> If data is cached with the probabibility of Ps(k), then data would not be
> likely to be cached where it is not likely to be found.  This would also
We don't want to use raw Ps(k). It's REALLY low (less than 0.1 except in
cases of flukes due to not very much data), look at your probability
histograms. Maybe Ps(k)/average Ps, capped at some fixed number - 1.0 or
0.9something, say.
> decrease I/O to the hard drive and increase node specialization.  A function
> more specific to probabalistic caching could be used, such as Pc(k) =
> probability of caching ~= 1 - abs(k - average(all keys in datastore))/(max
> key value), which would tend to get nodes to tend to specialize more toward
> one section of the keyspace.
What if the node has multiple specialities, as they normally do?
> Any thoughts?
> 
> 
> Scott Young

Attachment: msg03352/pgp00000.pgp
Description: PGP signature

Reply via email to