On Tue, Dec 09, 2008 at 12:43:15AM -0500, Daryl C. W. O'Shea wrote: > The long standing bottleneck (for net-enabled mass-checks) in my > otherwise fast mass-check cluster is Bind. It seems that it simply > cannot handle the load of a dozen or so cores worth of mass-check > processes. 475 kmsgs/hr non-net versus 70 kmsgs/hr net-enabled is > unacceptable.
I'm not saying that bind is great, but if you are telling us that it can only process 200 requests per second (10 requests per msg?), then something is definately wrong on your end. Even my 350Mhz router can handle 1000 reqs/ps!
