Shared memory and IPC

[Rich Salz]

> absolutely love to.  however, it's a speed improvement of
> a potential two or maybe even three orders of magnitude,

emphasize potential.

and of course, the syscall thrashing overhead of finding a common
mappable address may just kill you.  it'd be nice to see some kinds of
tests.

> the local bit, suddenly things like packing a 1mb data
> structure into a buffer, only to pass it over with a
> memcpy or five over to another program on the same host,
> become significant.

i think that would be the thing to work on for the biggest gain;
in-place marshalling.
once you can do that, use loopback networking might not be a bottleneck.

> the same way that those DB apps are doing, you must
> preserve pointers / memory segment offsets.

not sure SQL supports pointers. :)

> if anyone has any better ideas, love to hear them.

1.  profile.
2.  make the other stuff go faster first.
3.  don't forget about overall system load.
	/r$
-- 
Zolera Systems, Securing web services (XML, SOAP, Signatures,
Encryption)
http://www.zolera.com