directory-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Serge Knystautas <ser...@lokitech.com>
Subject Re: [MINA] Blocking I/O
Date Sun, 24 Apr 2005 06:57:50 GMT
Noel J. Bergman wrote:
> The following article:
> 
>   http://www-106.ibm.com/developerworks/library/j-nioserver/
> 
> demonstrates how to code the Servlet API using NIO.  Not only does it
> demonstrate why Tomcat needs to be updated to NIO, but it demonstrates how
> to service blocking I/O code using NIO.

Quoting from the article, "Some advanced app servers are optimized with 
JNI native code to improve scalability and therefore don't offer a good 
comparison between standard I/O and NIO" was probably referring to Resin...

http://www.caucho.com/resin-3.0/performance/faq.xtp

When they rebuilt Resin for the 3.x release, they compared existing Java 
IO, Java NIO, and a JNI code.  They found the gain from NIO was not as 
significant as just using a more efficient Java socket impl (using JNI).

That said, a protocol such as SMTP and POP3 would benefit much more from 
NIO than HTTP, since aside for keepalives, HTTP is designed to not have 
connections sitting idle or back-and-forth communication.  In fact, 
that's the beauty of HTTP. :)

-- 
Serge Knystautas
Lokitech >> software . strategy . design >> http://www.lokitech.com
p. 301.656.5501
e. sergek@lokitech.com

Mime
View raw message