Re: [Jetty-support] Jetty benchmark

[Greg W. <gr...@mo...>]

Yoni,

What are you using for a client?
Note that response time and throughput are not directly related and that the response
time for 1 request on 1 connection is not a good measure.

cheers


Yoni Moses wrote:
> Hi,
> i'm trying to do a small benchmark with jetty 4.2.12 and i'm getting REALLY bad results... 
> i'm guessing i'm doing something wrong, so perhaprs you can help me out here..
> my configuration is 
> 
>        HttpServer httpServer = new HttpServer();
>        SocketListener socketListener= new SocketListener();
>         socketListener.setPort(9090);
>         socketListener.setMinThreads(30);
>         socketListener.setMaxThreads(300);
>         socketListener.setLingerTimeSecs(-1);
>         //socketListener.setMaxIdleTimeMs();        
>         socketListener.setHost("yonim.dotomi.com");
>         httpServer.addListener(socketListener);
>         HttpContext httpContext = new HttpContext(httpServer,"/");
>         httpContext.setResourceBase("/home/dotomi/test_jetty");
>         httpContext.addHandler(new ResourceHandler());
>         httpContext.addHandler(new NotFoundHandler());
>         httpContext.start();
>         socketListener.start();
>         httpServer.start();
> 
> I've been playing with the setMaxIdleTimeMs from 10 to 5000 and the linger time on and off...
> the best results i came up was 40 millis to a servlet and 25 millis to a static page (2k page).
> Apache 2 serves the same page in 7 millis.
> resin serves the same page in 9-11 millis..
> 
> i'm aware that jetty is very fast, so am i doing wrong :-( ????
> 
> any help will be appreciated,
> Yehonatan Mozes.
> 
> 
> 
> 
> 
> 
> 
> -------------------------------------------------------
> This sf.net email is sponsored by:ThinkGeek
> Welcome to geek heaven.
> http://thinkgeek.com/sf
> _______________________________________________
> Jetty-support mailing list
> Jet...@li...
> https://lists.sourceforge.net/lists/listinfo/jetty-support
>