http://users.livejournal.com/_slw/ ([identity profile] http://users.livejournal.com/_slw/) wrote in [personal profile] nuclight 2011-08-04 06:56 am (UTC)

>э, а нахрена?

Ответ комплексный.

1)http://alouche.net/blog/2010/08/24/rps-and-rfs-kernel-network-stack/
2)http://kerneltrap.org/mailarchive/linux-netdev/2010/7/22/6281568

"Some legacy applications being not SMP friendly, one way to scale a
server is to run multiple copies of them.

Instead of randomly choosing an instance, we can use the cpu number as a
key so that softirq handler for a whole instance is running on a single
cpu, maximizing cache effects in TCP/UDP stacks."

Как пример таких приложений - node.js, python-gevent, bind.

не вижу ответа.
это дело шедулера знать и учитывать.
зачем это изнутри pf знать?

Post a comment in response:

This account has disabled anonymous posting.
If you don't have an account you can create one now.
HTML doesn't work in the subject.
More info about formatting