Brian Aker brian at
Sun Dec 2 21:35:51 UTC 2007


On Nov 30, 2007, at 8:31 AM, Alberto Bertogli wrote:

> Or nmdb (, which can also handle  
> the
> "store the same data in multiple servers, just in case one fails"

libmemcached has a consistent hash it uses to allocate data out to  
servers. The "replica" code will be pushed soon. It will store data  
into additional nodes in a way that you can lose a node, or add a note  
without a significant, or possibly any, loss of cache.


Brian "Krow" Aker, brian at
Seattle, Washington                     <-- Me                <-- Software    <-- Fun
You can't grep a dead tree.

More information about the memcached mailing list