extremely high context switches and interrrupts

[Jon Gabrielson]

I converted my existing database from myisam to a ndb cluster.
In doing so my benchmark went from 14 seconds to 2 minutes.
CPU, disk, and network usage all seem relatively low, but
my context switches and interrupts are both above 40k on both
the mysql node and the ndb nodes, so I'm assuming that is where
the bottleneck is. Does anyone know a way to reduce the number
of interrupts and context switches?


Thanks,


Jon.