I'm actually pretty fuzzy about it myself. Say handling nondatabase tasks
means the throughput for handling pure http connection serving capabilities,
ie., x http connections per second. And database tasks
means the throughput for handling http connections plus database activities
(assuming average data retrieval of 1Kb), ie, x http connections + y
database accesses per second. For the later case, i'm not sure whether jdbc
speed will be the bottleneck. Plus, there's a possibility that the number
of sessions concurrently maintained might impose a limit on how many
sessions are maintenable at the same time.
In real life for a large scale project(200+ concurrent sessions and
connections, 100+ database accesses), what the real bottleneck will be and
what hardware and software technology can be used to deal with them?
Thanks,
Zheng
>From: Michal Mosiewicz <[EMAIL PROTECTED]>
>Reply-To: "A mailing list for discussion about Sun Microsystem's Java
> Servlet API Technology." <[EMAIL PROTECTED]>
>To: [EMAIL PROTECTED]
>Subject: Re: Performance Jsp vs. Asp vs. cgi-perl vs. Php vs. mod-perl
>Date: Mon, 22 Nov 1999 00:23:17 +0100
>
>zheng li wrote:
> >
> > Hi,
> >
> > I could not find any performance comparisons on net. Do any one know
>how
> > does Jsp compare vs. Asp vs. cgi-perl vs. Php vs. mod-perl for doing
> > following tasks:
> > 1. database access assuming mySQL
> > 2. non database
>
>How do you want to benchmark it? It's not that easy. As a case study - I
>moved one of my projects from mod-perl to JSDK. The performance
>increased very drastically. Why?
>
>1. I would need about 600MBs of memory in my server, while now 128MB is
>more than sufficient.
>2. I can integrate web applications with external apps in a single
>process which gives me a lot of features and I don't have to use
>frequent resource checks - I can use better mechanisms like
>notifications and direct modifications of internal data.
>3. I can share a lot of useful data between concurrent and successive
>requests.
>4. Becouse of (3) I was able to reduce database load 3 times.
>
>Then - how can you compare which language is faster? Some problems
>implemented in Java gave me 4000% better performance than in Perl due to
>availability of better data model, better structures, strictier typing,
>etc.
>
>So, again - what is the correct meaning of database vs non-database
>tasks? I mean - in terms of benchmarks.
>
>-- Mike
>
>___________________________________________________________________________
>To unsubscribe, send email to [EMAIL PROTECTED] and include in the body
>of the message "signoff SERVLET-INTEREST".
>
>Archives: http://archives.java.sun.com/archives/servlet-interest.html
>Resources: http://java.sun.com/products/servlet/external-resources.html
>LISTSERV Help: http://www.lsoft.com/manuals/user/user.html
______________________________________________________
Get Your Private, Free Email at http://www.hotmail.com
___________________________________________________________________________
To unsubscribe, send email to [EMAIL PROTECTED] and include in the body
of the message "signoff SERVLET-INTEREST".
Archives: http://archives.java.sun.com/archives/servlet-interest.html
Resources: http://java.sun.com/products/servlet/external-resources.html
LISTSERV Help: http://www.lsoft.com/manuals/user/user.html
