Thanks for the reply. I didn't realize you could chose your own connector
with JRun. But I have another question or two: why would your company's
Apache solution be more stable than an in-process solution? Is this a
single-point-of failure issue, e.g., not wanting a problem in the servlet
engine to take down the entire Web site?
Even though you say it's obvious, I'm not sure what a stand-alone is (You
still need a Web server, don't you?), nor why it is inherently
inferior. Please explain what this is.
Thanks,
Bruce
At 12:49 AM 8/27/00 -0400, you wrote:
>Hi Bruce,
>Over the past year I've been working with and testing servlet engines and
>configurations both for my home based web business and my employer. Here's
>the info you need:
>You're choices are in-process, as with Netscape server, out-of-process, as
>with Apache and mod_jserv, and stand-alone. I think the reasons for not using
>a stand-alone servlet engine are obvious. In-process vs out-of-process is a
>trade off between speed and stability. For my little home based web business
>where speed is a greater issue than stability I use an in-process servlet
>engine (JavaWebServer). But, for my employer where stabilty is as great an
>issue as speed we're using Apache with Jrun and the mod_jserv connector.
>
>Bruce Altner <[EMAIL PROTECTED]> wrote:
>
> > ---------------------------------------------
> > Attachment:
> > MIME Type: multipart/alternative
> > ---------------------------------------------
> > Greetings:
> >
> > I'm part of a team working on a project to evaluate performance issues
> > concerning implementing servlets and JSPs for our customer, which is a big
> > government agency. The end result will be a series of recommendations to go
>
> > or not to go in this direction and, if it's a go (as appears likely), to
> > also recommend implementation strategies.
> >
> > Every time the subject of JRun as a servlet engine comes up at our meetings
>
> > someone voices the opinion that because it's an add-on to the web server,
> > via the connector, instead an integral part of the server itself (e.g., as
> > in Netscape server), that it would be slower. While this sounds reasonable,
>
> > I was wondering if there is any data to back this up. Does anyone know of
> > studies done to evaluate the performance of application server/servlet
> > engine products and whether these indeed show that plug-in engines are any
> > slower than those already built into the server?
> >
> > Actually, I'm interested in detailed performance evaluations/comparisons
> > involving any aspect of server-side Java and would appreciate leads there
>too.
> >
> >
> > Thanks in advance,
> > Bruce Altner
>
>
>____________________________________________________________________
>Get free email and a permanent address at http://www.netaddress.com/?N=1
>------------------------------------------------------------------------------
>Archives: http://www.egroups.com/group/jrun-interest/
>Unsubscribe:
>http://www.houseoffusion.com/index.cfm?sidebar�sts&body�sts/jrun_talk
>or send a message to [EMAIL PROTECTED] with
>'unsubscribe' in the body.
------------------------------------------------------------------------------
Archives: http://www.egroups.com/group/jrun-interest/
Unsubscribe: http://www.houseoffusion.com/index.cfm?sidebarRsts&bodyRsts/jrun_talk
or send a message to [EMAIL PROTECTED] with 'unsubscribe' in the
body.
