Any news on this (sorry if I missed it)?
Are both hosts solaris ?
The messages are single packets ? (Otherwise we'd check for
Nagle issue).
I see three timers in this order of magnitude.
By tuning each one individually with ndd we would try to get
a picture for which one influences the result.
On Tue 03 Oct 2006 at 04:01PM, Eric Saxe wrote:
> >Timing measurements were made with snoop.
I just saw a mail from Jim Carlson over on dtrace-discuss about how snoop's
timestamps aren't very accurate due to its use of bufmod...
http://www.opensolaris.org/jive/thread.jspa?messageID=61494
I'm
Hi Jim,
Jim M wrote:
Hello.
One of the developers I support is having a performance challenge. He is writing a simple message processing application that receives a small message on a socket and writes a response to that socket, then waits for the next message. The application runs one thread
Hello.
One of the developers I support is having a performance challenge. He is
writing a simple message processing application that receives a small message
on a socket and writes a response to that socket, then waits for the next
message. The application runs one thread per client connectio
