That is correct, this is a relevant fix for improved performance on most server machines, especially with multi-core quickly becoming ubiquitous. It wouldn't be hard to change the default constructor. Perhaps you could open a JIRA issue to request this, or submit a small patch, as MINA developers have a lot to track right now. If there are other sensible defaults to maximize out-of-the-box performance, we should include those as well.
On 3/31/07, James Im <[EMAIL PROTECTED]> wrote:
Ok, so you seems to have missed a 'go fast' parameter. In my opinion, it is not your fault. It is mina that is not being user friendly. MINA should take care of these details itself. This optimization wisdom should be capture into mina so that everyone is not forced to be an expert to have a fast server. For me, mina should be fast by default (if you want to overwrite the config you can). Right now mina is slow by default and only if you tweak some settings it becomes faster. This is not right. Emmanuel Lecharny wrote: > >Greg, > >I have done some more test, on the 'big server' ( 8 CPUs), changing the >SocketAcceptor initialization with something like what you suggested : > > tcpAcceptor = new SocketAcceptor( > Runtime.getRuntime().availableProcessors(), >threadPoolExecutor ); > >(we already have this threadPoolExecutor, so I reuse it). > >The result is huge : I can get as much as 6400 request/s, instead of >only 2000 previously ! And the CPU reach 85% of user+syst usage (70% >user, 10 to 15% system) > >This is just what I expected :) > >I have more tests to do, but this is a big step in the good direction ! > >Thanks a lot ! > >Emmanuel > _________________________________________________________________ Ta' på udsalg året rundt på MSN Shopping: http://shopping.msn.dk - her finder du altid de bedste priser
