hc-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Robert Olofsson <r...@khelekore.org>
Subject Re: [HttpCore] NIO extensions: progress report
Date Wed, 30 Aug 2006 20:32:06 GMT
Tatu Saloranta wrote:
> That is one aspect, but to me scalability aspects
> regarding threading efficiency are more interesting.
> When serving thousands of active (but not
> hyperactive!) clients there will be a point where
> network still allows more traffic to go through, but
> the server overhead due to thread-management kills
> performance, at least when using 'classic' one thread
> owns the transaction paradigm (like servlet API).

I know, my web proxy is aimed at speeding up slow
connections, that means modem pools and similar things.
The old version did use one thread per client and it
had problems with many (not sure exactly > 1000) clients.
The newer, nio based, versions seems better.
Considering that modem users are very slow nio seems very good.

> streaming operation. But it's really the client-side
> functionality where I haven't seen combination where
> number of worker threads can scale somewhat
> independent of number of (logical) connections being
> used.

I have seen client side functionality, but that may be an exception
since I did write it myself...

You can look at my web proxy, rabbit it does this.
http://www.khelekore.org/rabbit/

Since it is a web proxy it handles both client and server side.
It does not do http they way HttpClient does it, it is
on a lower level, but it may still be an interesting thing
to look at.

> Think of a case where you have to serve 1000 requests
> per second, ... 

I have done that already, see above.. :-)

/robo

---------------------------------------------------------------------
To unsubscribe, e-mail: httpclient-dev-unsubscribe@jakarta.apache.org
For additional commands, e-mail: httpclient-dev-help@jakarta.apache.org


Mime
View raw message