Re: Java.exe Application Error

[GCS]

Hi!

On H, NOV 20, 2000 at 11:27:30 -0700, Sam T.Stanley wrote:

>  Application Error and the tomcat console is closed on clicking "ok"
 I used the same jdk and tomcat version under Linux. I got the same problem,
and I reported it to SUN. Unfortunately it was a too quick report, without
getting deep. So they can't find out, and I have got the new jdk since then.
It works with that. So, get jdk1.3.
Bye,
Laszlo

Antwort: Re: Antwort: RE: tomcat vs orion vs weblogic testing results (LoadRunner)

[thomas . maesing]

Yes,
it would be very nice to get the complete architecture for these results.

Thomas





"Miles Daffin" <[EMAIL PROTECTED]> am 21.11.2000 00:46:37

Bitte antworten an [EMAIL PROTECTED]

An:<[EMAIL PROTECTED]>
Kopie:
Thema: Re: Antwort: RE: tomcat vs orion vs weblogic testing results (Load
   Runner)


> Oops. Tomcat twice the speed of WebSphere.

Which versions/builds of WebSphere and TomCat would these be exactly?
Can you say anything else - some settings, stats etc.

Miles

SnoopServlet - passing params through web.xml

[Mala Ramakrishnan]

Hi,

I wanted to learn how to pass init parameters through web.xml, so I
uncommented the following from
jakarta-tomcat-3.1/webapps/examples/WEB-INF/web.xml:

foo
   bar


Yet, when I call SnoopServlet on my browser as follows:
http://www.blahblah.com:8080/examples/servlet/SnoopServlet

I get empty fields in Servlet init parameters.

I tried the same with rewriting my own SnoopServlet and yet get null in
this field. Why would that be?

Mala

Re: Questions re: Tomcat and IBM VAJ

[Olivier Baillard]

Hello,

http://www7.software.ibm.com/vad.nsf/Data/Document2390?OpenDocument&p=1&BCT=3&Footer=1


Bye!
Olivier Baillard



At 16:58 20.11.00 -0500, you wrote:
>Hello. I'm trying to get Tomcat running under Windows 'NT with IBM VAJ.
>Unfortunately, I haven't been able to make everything work together just yet,
>and would greatly appreciate your help in resolving what is probably a very
>simple problem.
>
>Environment
>  - MS-Windows 'NT 4.0
>  - IBM VAJ version 3.0 for J2SE 1.2 (the free, up to 750 classes version of
>VAJ)
>  - Tomcat 3.1
>
>All of the above have been downloaded and installed into their "default"
>locations. Environment variables include the following:
>CLASSPATH=C:\IBMVJava2\eab\runtime30
>JAVA_HOME=C:\IBMVJava2\eab\runtime30
>TOMCAT_HOME=C:\jakarta\tomcat
>
>When I enter the following commands:
>C:\jakarta\tomcat\bin>startup
>Starting tomcat in new window
>Using classpath:
>c:\jakarta\tomcat\classes;c:\jakarta\tomcat\lib\webserver.jar;c
>:\jakarta\tomcat\lib\jasper.jar;c:\jakarta\tomcat\lib\xml.jar;c:\jakarta\tomca
>
>t\
>lib\servlet.jar;c:\j2sdkee1.2.1\lib\tools.jar;C:\IBMVJava2\eab\runtime30;
>C:\jakarta\tomcat\bin>
>
>A new MS-DOS Command Prompt window opens with the title
>"c:\WINNT\system32\java.exe". It contains the following output, presumably
>from Tomcat:
>Context log: path="/examples" Adding context path="/examples"
>docBase="webapps/
>examples"
>Context log: path="" Adding context path=""  docBase="webapps/ROOT"
>Context log: path="/test" Adding context path="/test"  docBase="webapps/test"
>Starting tomcat. Check logs/tomcat.log for error messages
>Starting tomcat install="c:\jakarta\tomcat" home="C:\jakarta\tomcat"
>classPath="
>c:\jakarta\tomcat\classes;c:\jakarta\tomcat\lib\webserver.jar;c:\jakarta\tomca
>
>t\
>lib\jasper.jar;c:\jakarta\tomcat\lib\xml.jar;c:\jakarta\tomcat\lib\servlet.jar
>
>;c
>:\j2sdkee1.2.1\lib\tools.jar;C:\IBMVJava2\eab\runtime30;"
>Context log: path="/admin" Automatic context load
>docBase="C:\jakarta\tomcat\web
>apps\admin"
>Context log: path="/admin" Adding context path="/admin"
>docBase="C:\jakarta\tom
>cat\webapps\admin"
>
>The logs do not seem to indicate an error message -- the tomcat.log contains
>the following:
>Starting endpoint port="8080"
>handler="org.apache.tomcat.service.http.HttpConnectionHandler" Starting
>endpoint port="8007"
>handler="org.apache.tomcat.service.connector.Ajp12ConnectionHandler"
>
>
>While this seems to indicate that Tomcat is working properly, I know it
>isn't. I have two files, simple.html and simple.jsp (downloaded from Wrox, as
>I'm using their Professional JSP book as a reference).
>begin simple.html 
>
>
>Simple JSP Example
>
>
>
>
>How many times?
>
>
>
>
>
>
>
>
> end simple.html 
> begin simple.jsp 
><%@ page language="java" %>
>
>
>
>Simple JSP Example
>
>
>
>
>
>   <% int numTimes = Integer.parseInt(request.getParameter("numtimes"));
>  for (int i = 0; i < numTimes; i++) {
>   %>
>Hello, world!
>   <% }
>   %>
>
>
><%@ include file="PageFooter.html" %>
>
>
>
> end simple.jsp 
>
>simple.html loads fine. When I enter a number into the resultant form
>simple.jsp is loaded but the JSP servlet code IS NOT executed. What am I
>doing wrong?
>
>Luke Hohmann

Java.exe Application Error

[Sam T . Stanley]

Hi,
I am using IIS with Tomcat in NT Server.Everthing
 works fine(?) except it results in "java.exe"
 Application Error and the tomcat console is closed on clicking "ok"
 please suggest me a solution
 I am using html,jsp,servlet in this web application.
 Also using sql server 6.5 with odbc bridge.

jdk1.2.2
Tomcat3.1
IIS 4.0 
NT pack 6

please help.

Stanley





Get free email and a permanent address at http://www.netaddress.com/?N=1

RE: Would you please post your config files

[YUTUNG93]

Title: RE: Would you please post your config files





Hi,
If you got the answers to the following questions, could you kindly send me a copy? 'cause I have the same questions. Thanks.

Regards,
Tony


-Original Message-
From: Debbie Cao [mailto:[EMAIL PROTECTED]]
Sent: Tuesday, November 21, 2000 6:06 AM
To: '[EMAIL PROTECTED]'
Subject: Would you please post your config files



Hi, Larry,
I am configuring Apache, Tomcat on NT 4.0 Workstation. The configuration
is really a pain. Your message helped a lot. There are still a few questions


not very clear to me:
1. The port numbers. Where should I say 8080, where to say 80, 8007,
8040, localhost, ...
2. How didyou configure Tomcat using Ajpv12 protocol?
3. JServ has the concept of "zone". Does Tomcat support this feature?
So, I am very much interested in reading your config files. Would you please
post your configu files. Or, if you don't mind, please send them to:
    [EMAIL PROTECTED]
or  [EMAIL PROTECTED]
Thanks a lot,
Dehe Cao

"Bug"

[Vinayak Infotech]

Hello,
 
We are developing an application in JSP and 
Servlets and SQL 6.5 as back end using Tomcat 3.1 as the servlet engine in IIS 
4.0(Service Pack6.0). We have developed the application with JDK 
1.2.2.  and connecting with SQL using JDBCODBC driver.  The 
application is used to generate a chart on entering the 
staistics.  But when we input the data nd try to generte the 
chart "Java.exe Error" occurs and the servlet engine gets closed.  The 
chart is displayed in th back ground.
 
Please find attached the screen shot of the error 
report.
 
-Kamal. 
 Error.jpg

RE: One user -> 2 parallel threads

[Aaron Knauf]

Well, that's not entirely true.

'isSingleThreaded'  sets the generated servlet to implement SingleThreadModel.  This does not actually cause subsequent requests to be queued.  It causes another instance of the servlet to be instantiated for each simultaneous request.

This is generally considered to be 'a bad thing' (tm).

The only thing that is singly threaded in this is the servlet class itself.  All other objects accessed  (e.g. session/context objects) are accessed in a multithreaded manner.

I suggest that your client side fix might involve using the onDblClick javascript event handler (possibly configured to simply do nothing).


Aaron Knauf
Systems Integrator
Genie Systems Ltd
Auckland, New Zealand
Ph. +64-9-573 3310 x812
email: [EMAIL PROTECTED]
http://www.geniesystems.com







Steve Toth <[EMAIL PROTECTED]>
21/11/2000 06:06
Please respond to tomcat-user

        
        To:        "'[EMAIL PROTECTED]'" <[EMAIL PROTECTED]>
        cc:        
        Subject:        RE: One user -> 2 parallel threads


You could also make the jsp singlethreaded.

<%@ page isSingleThreaded="false" %>

This will tell tomcat that the jsp should only be in one thread, multiple
requests will be queued up. thus the jsp should be FAST (no IO/DB access).
Same as implementing the servlet Interface SingltThreadModel.

 -Original Message-
From: Till Gartner [mailto:[EMAIL PROTECTED]]
Sent: Saturday, November 18, 2000 7:32 AM
To: '[EMAIL PROTECTED]'
Subject: AW: One user -> 2 parallel threads


You're right: We checked the Apache log and the user just double clicked.

But Netscape seems to have a bug that cannot be handled with the JavaScript
solution:
If you have your security settings in a way that he comes up with a
messagebox before sending out POSTs, he sometimes "overhears" the first
click in that box and only reacts on the second click. Strangely he then
also sends out 2 requests.

How does the JavaScript to catch this kind of doble clicks look like?

Thanks for your help ona rainy saturday,
-- Till.

-Ursprüngliche Nachricht-
Von: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED]]
Gesendet: Samstag, 18. November 2000 16:21
An: [EMAIL PROTECTED]
Betreff: Re: One user -> 2 parallel threads


So, in those cases, I've built JavaScript logic running in the browser (in
the HTML file) that will set a boolean flag to block the second (and
subsequent) click(s).  Is there a simple way to handle this server-side (I
can imagine the hard ways!), or does anyone know that the client side still
has holes I have not witnessed?

JT

<< Undoubtedly the user clicked twice on the button or link and
 actually created two requests.>>

 >
 > Hi mail list,
 >
 > we encounter the following problem with Tomcat. From time to time a
browser
 > request seems to be handled by two threads from tomcat in parallel. Our
own
 > log from java servlets shows entries of different threads doing the same
 > task for the same user, wich are just some milli secs apart. >>

Re: Tomcat StartUp Problem

[Jim Rudnicki]

> Can somebody please help me solve the error I am getting  I try to start

> "Java.lang.ClassNot  FoundException:
... snip ...

Try this Tomcat loader.  Works 99.44% of the time.  Take a look:
http://www.geocities.com/jdrudnicki/

Jim

Help Needed

[Ramesh Kumar . T]

Hi ,

I had previously seen some
xml tags in earlier versions of tomcat (I do not remember the version)
something like


(but this does not work right now )

instead of the
<%
%>
tags

  Is the Xml  format going to be a standard or some
thing?
can somebody  exaplain this

I would appreciate any links to Model2 architecture  described by sun
and links for writing tag libraries

One more thing I just tried Tomcat4.0 beta .
What is this catalina???
I Could not find any documents describing  the architecture of tomcat .
Can anyone give me some pointers???

Thanks
Ramesh

Tomcat and UltraDev

[Andrea Bertone]

Is anyone out there using 
UltraDev to develop jsp pages. I am trying to set the run-time connection to no 
avail.. I have Tomcat installed on my Linux server and have a database built 
with MySQL.
 
Please, can someone help... I 
wanna cry.
 
 
Andrea Bertone
www.bertoneadvertising.com
ph. 562.404.7300
fax 562.623.0027
 

Re: JSP 1.1 XML format tags

[Anne and Bill Moore]

"Craig R. McClanahan" wrote:
> 
> Anne and Bill Moore wrote:
> 
> > Does Tomcat support JSPs where the tags are coded in XML format as
> > specified in the JSP1.1 spec? I have tried Tomcat V3.1 and 3.2 beta 7
> > and they do not seem to recognize XML format tags.
> 
> For Tomcat 3.x (JSP 1.1), the answer is no.
> 
> For Tomcat 4.0 (JSP 1.2), the answer is yes.
> 
--
Craig
Thanks for the clarification - it seems I was working from an old
version of the JSP1.1 spec or had not remebered it correctly - I just
checked the final release again and it has the following all important
sentences about XML tags:
"The JSP page to XML document mapping is not visible to JSP 1.1
containers; it will receive substantial emphasis in the next releases of
the JSP specification. Since the mapping has not received great usage,
we particularly encourage feedback in this area."

I am sure you knew this anyway - it certainly explains why its not in
Tomcat 3.x, but thanks again for the reply.
 
Thanks and Regards
Anne and Bill Moore
[EMAIL PROTECTED]
Remember: Life is too short to drink bad wine!!!

[ANNOUNCEMENT] Tomcat 3.2 Beta 8

[Craig R. McClanahan]

Thanks to everyone for the bug reports (and patches)!  You can now download
"beta 8" of the Tomcat 3.2 servlet container from:

http://jakarta.apache.org/builds/tomcat/release/v3.2-beta-8

Relative to "beta 7", a number of critical bugs have been fixed, and several
configuration options available under Tomcat 3.1 have been restored (although
they are done in "conf/server.xml" rather than "conf/web.xml").

Please download this release and check it out.  If we've squashed all the
critical bugs, we can proceed to a final release!  Non-critical bugs will be
noted in the release notes and can be fixed in a later maintenance cycle.

NOTE:  No changes were made in any of the native code modules since version
"beta 7".  There are a number of outstanding bug reports against MOD_JK and
AJP13 that will remain open in the final release unless someone volunteers to
find and fix them.

Craig McClanahan

Re: JSP 1.1 XML format tags

[Craig R. McClanahan]

Anne and Bill Moore wrote:

> Does Tomcat support JSPs where the tags are coded in XML format as
> specified in the JSP1.1 spec? I have tried Tomcat V3.1 and 3.2 beta 7
> and they do not seem to recognize XML format tags.

For Tomcat 3.x (JSP 1.1), the answer is no.

For Tomcat 4.0 (JSP 1.2), the answer is yes.

> --
> Thanks and Regards
> Anne and Bill Moore
> [EMAIL PROTECTED]
> Remember: Life is too short to drink bad wine!!!

or beer :-)

Craig McClanahan

JSP 1.1 XML format tags

[Anne and Bill Moore]

Does Tomcat support JSPs where the tags are coded in XML format as
specified in the JSP1.1 spec? I have tried Tomcat V3.1 and 3.2 beta 7
and they do not seem to recognize XML format tags.
-- 
Thanks and Regards
Anne and Bill Moore
[EMAIL PROTECTED]
Remember: Life is too short to drink bad wine!!!

Re: Portability & other issues of the minute.

[Craig R. McClanahan]

See intermixed.

Miles Daffin wrote:

> Hello to all those treading the Tomcat user trail,
>
> 1: Fully encapsulated webapps...
>
> > >
> > > Because we're trying to write nice well-behaved webapps, we don't want
> to
> > > have to change a global setting, only a local-to-webapp setting.
> > >
>
> Is it possible to set the access control stuff on a webapp local scope? I am
> talking about the stuff currently defined in 'tomcat-users.xml'.
>

In Tomcat 3.2, no.  In Tomcat 4.0, yes.

>
> It would be better in my opinion.
>
> 2: Load on start up - funny numbers
> e.g.
>   
> -2147483646
>   
>
> Why -2147483646 here (taken from the servlet entry in main web.xml for
> org.apache.tomcat.servlets.DefaultServlet)
> and 5 in the example web.xml in the doc\appdev directory?
>

It's the most negative integer that can be represented in a 32-bit int.
Essentially, this value (or any negative value) says "I want this servlet loaded
at startup time, but I do not care what order it is loaded in."

>
> 3: How can I use ANT to compile my JSPs so I can debugger them without
> having to go through TOMCAT??
> This would be a really useful trick. I have been trying in vain to get
> Jasper to do this.
>

One approach would be to use the "jspc" command-line tool (which you can invoke
from Ant using the  task).  This will precompile all of the JSP pages in
your app, and give you the source code for them (plus a chunk of stuff to stick
in web.xml for the generated servlets).

Of course, you still need Tomcat to run the JSP pages while debugging them,
because JSP pages are turned into servlets (not standalone applications).  Some
IDEs (like Forte) do this for you behind the scenes.

>
> Miles

Craig

Jasper

[Miles Daffin]

Hi,
 
I have managed to invoke Jasper from a batch file. 

 
Either I get a list of options because the JSP file 
is not found
or Japser is mysteriously quiet - no news, no 
apparent compilation.
 
What am I doing wrong?
 
Miles

Re: Returned mail: User unknown

[Miles Daffin]

> Yeah...  the goofball slempine used the wrong email address when he/she
> registered... 
> 
> So, unless we create a filter they will keep showing up.
> 

That mesage was so garbled - which address do I filter out?

Miles

Portability & other issues of the minute.

[Miles Daffin]

Hello to all those treading the Tomcat user trail,

1: Fully encapsulated webapps...

> >
> > Because we're trying to write nice well-behaved webapps, we don't want
to
> > have to change a global setting, only a local-to-webapp setting.
> >

Is it possible to set the access control stuff on a webapp local scope? I am
talking about the stuff currently defined in 'tomcat-users.xml'.

It would be better in my opinion.

2: Load on start up - funny numbers
e.g.
  
-2147483646
  

Why -2147483646 here (taken from the servlet entry in main web.xml for
org.apache.tomcat.servlets.DefaultServlet)
and 5 in the example web.xml in the doc\appdev directory?

3: How can I use ANT to compile my JSPs so I can debugger them without
having to go through TOMCAT??
This would be a really useful trick. I have been trying in vain to get
Jasper to do this.


Miles

RE: Returned mail: User unknown

[Mark Franz]

Yeah...  the goofball slempine used the wrong email address when he/she
registered... 

So, unless we create a filter they will keep showing up.

Mark G. Franz


-Original Message-
From: Miles Daffin [mailto:[EMAIL PROTECTED]]
Sent: Monday, November 20, 2000 4:03 PM
To: [EMAIL PROTECTED]
Subject: Fw: Returned mail: User unknown


Do you all keep getting these?

Why?

Can we stop em?


- Original Message -
From: "Mail Delivery Subsystem" <[EMAIL PROTECTED]>
To: <[EMAIL PROTECTED]>
Sent: Tuesday, November 21, 2000 2:00 AM
Subject: Returned mail: User unknown


> The original message was received at Tue, 21 Nov 2000 03:00:00 +0200
> from IDENT:[EMAIL PROTECTED] [127.0.0.1]
>
>- The following addresses had permanent fatal errors -
> 
>
>- Transcript of session follows -
> 550 ... User unknown
>






> Return-Path: <[EMAIL PROTECTED]>
> Received: from localhost (IDENT:[EMAIL PROTECTED]
[127.0.0.1])
> by octarine.ntc.nokia.com (8.9.3/8.9.3) with ESMTP id DAA21913
> for ; Tue, 21 Nov 2000 03:00:00 +0200
> Received: from treis02nok.europe.nokia.com
> by localhost with IMAP (fetchmail-5.3.1)
> for slempine@localhost (single-drop); Tue, 21 Nov 2000 03:00:00 +0200
(EET)
> Received: from mgw-x1.nokia.com ([131.228.20.21]) by
esebh02nok.ntc.nokia.com with SMTP (Microsoft Exchange Internet Mail Service
Version 5.5.2652.78)
> id XCQWC5L4; Tue, 21 Nov 2000 01:54:43 +0200
> Received: (from root@localhost)
> by mgw-x1.nokia.com (8.10.2/8.10.2/Nokia) id eAKNsge01549
> for [EMAIL PROTECTED]; Tue, 21 Nov 2000 01:54:42 +0200 (EET)
> Received: from locus.apache.org (locus.apache.org [63.211.145.10])
> by mgw-x1.nokia.com (8.10.2/8.10.2/Nokia) with SMTP id eAKNsfK01539
> for <[EMAIL PROTECTED]>; Tue, 21 Nov 2000 01:54:41 +0200 (EET)
> Received: (qmail 25892 invoked by uid 500); 20 Nov 2000 23:51:03 -
> Mailing-List: contact [EMAIL PROTECTED]; run by ezmlm
> Precedence: bulk
> Reply-To: [EMAIL PROTECTED]
> list-help: 
> list-unsubscribe: 
> list-post: 
> Delivered-To: mailing list [EMAIL PROTECTED]
> Received: (qmail 25724 invoked from network); 20 Nov 2000 23:50:59 -
> Received: from cable064.228.eneco.bart.nl (HELO hotlink.xs4all.nl)
(@195.38.228.64)
>   by locus.apache.org with SMTP; 20 Nov 2000 23:50:59 -
> Received: from hotduck ([10.0.2.222])
> by hotlink.xs4all.nl (8.9.3/8.9.3) with SMTP id AAA03730
> for <[EMAIL PROTECTED]>; Tue, 21 Nov 2000 00:49:09 +0100
> Message-ID: <007901c0534c$2a1234b0$de02000a@hotduck>
> From: "EXT Miles Daffin" <[EMAIL PROTECTED]>
> To: <[EMAIL PROTECTED]>
> References: <[EMAIL PROTECTED]>
> Subject: Re: Antwort: RE: tomcat vs orion vs weblogic testing results
(Load Runner)
> Date: Tue, 21 Nov 2000 00:46:37 +0100
> MIME-Version: 1.0
> Content-Type: text/plain;
> charset="iso-8859-2"
> Content-Transfer-Encoding: 7bit
> X-Priority: 3
> X-MSMail-Priority: Normal
> X-Mailer: Microsoft Outlook Express 5.50.4133.2400
> X-MimeOLE: Produced By Microsoft MimeOLE V5.50.4133.2400
> X-Spam-Rating: locus.apache.org 1.6.2 0/1000/N
>

Fw: Returned mail: User unknown

[Miles Daffin]

Do you all keep getting these?

Why?

Can we stop em?


- Original Message -
From: "Mail Delivery Subsystem" <[EMAIL PROTECTED]>
To: <[EMAIL PROTECTED]>
Sent: Tuesday, November 21, 2000 2:00 AM
Subject: Returned mail: User unknown


> The original message was received at Tue, 21 Nov 2000 03:00:00 +0200
> from IDENT:[EMAIL PROTECTED] [127.0.0.1]
>
>- The following addresses had permanent fatal errors -
> 
>
>- Transcript of session follows -
> 550 ... User unknown
>






> Return-Path: <[EMAIL PROTECTED]>
> Received: from localhost (IDENT:[EMAIL PROTECTED]
[127.0.0.1])
> by octarine.ntc.nokia.com (8.9.3/8.9.3) with ESMTP id DAA21913
> for ; Tue, 21 Nov 2000 03:00:00 +0200
> Received: from treis02nok.europe.nokia.com
> by localhost with IMAP (fetchmail-5.3.1)
> for slempine@localhost (single-drop); Tue, 21 Nov 2000 03:00:00 +0200
(EET)
> Received: from mgw-x1.nokia.com ([131.228.20.21]) by
esebh02nok.ntc.nokia.com with SMTP (Microsoft Exchange Internet Mail Service
Version 5.5.2652.78)
> id XCQWC5L4; Tue, 21 Nov 2000 01:54:43 +0200
> Received: (from root@localhost)
> by mgw-x1.nokia.com (8.10.2/8.10.2/Nokia) id eAKNsge01549
> for [EMAIL PROTECTED]; Tue, 21 Nov 2000 01:54:42 +0200 (EET)
> Received: from locus.apache.org (locus.apache.org [63.211.145.10])
> by mgw-x1.nokia.com (8.10.2/8.10.2/Nokia) with SMTP id eAKNsfK01539
> for <[EMAIL PROTECTED]>; Tue, 21 Nov 2000 01:54:41 +0200 (EET)
> Received: (qmail 25892 invoked by uid 500); 20 Nov 2000 23:51:03 -
> Mailing-List: contact [EMAIL PROTECTED]; run by ezmlm
> Precedence: bulk
> Reply-To: [EMAIL PROTECTED]
> list-help: 
> list-unsubscribe: 
> list-post: 
> Delivered-To: mailing list [EMAIL PROTECTED]
> Received: (qmail 25724 invoked from network); 20 Nov 2000 23:50:59 -
> Received: from cable064.228.eneco.bart.nl (HELO hotlink.xs4all.nl)
(@195.38.228.64)
>   by locus.apache.org with SMTP; 20 Nov 2000 23:50:59 -
> Received: from hotduck ([10.0.2.222])
> by hotlink.xs4all.nl (8.9.3/8.9.3) with SMTP id AAA03730
> for <[EMAIL PROTECTED]>; Tue, 21 Nov 2000 00:49:09 +0100
> Message-ID: <007901c0534c$2a1234b0$de02000a@hotduck>
> From: "EXT Miles Daffin" <[EMAIL PROTECTED]>
> To: <[EMAIL PROTECTED]>
> References: <[EMAIL PROTECTED]>
> Subject: Re: Antwort: RE: tomcat vs orion vs weblogic testing results
(Load Runner)
> Date: Tue, 21 Nov 2000 00:46:37 +0100
> MIME-Version: 1.0
> Content-Type: text/plain;
> charset="iso-8859-2"
> Content-Transfer-Encoding: 7bit
> X-Priority: 3
> X-MSMail-Priority: Normal
> X-Mailer: Microsoft Outlook Express 5.50.4133.2400
> X-MimeOLE: Produced By Microsoft MimeOLE V5.50.4133.2400
> X-Spam-Rating: locus.apache.org 1.6.2 0/1000/N
>

 ATT00178.dat

Re: Antwort: RE: tomcat vs orion vs weblogic testing results (Load Runner)

[Miles Daffin]

> Oops. Tomcat twice the speed of WebSphere.

Which versions/builds of WebSphere and TomCat would these be exactly?
Can you say anything else - some settings, stats etc.

Miles

Re: Session timeout

[Miles Daffin]

> > Note that you can also change the timeout on a
> > session dynamically, by calling
> > the session.setMaxInactiveInterval() method.
> >
> > > I v a n ...
> >
> > Craig McClanahan
>  but in that  tag what is the unit ?
> secondes, minutes?

This a classic "read the API docs and find out" answer.

If you goto
%TOMCAT_HOME\webapps\ROOT\docs\api\javax\servlethttp\HttpSession.html

you will find

 javax.servlet.http Interface HttpSession

setMaxInactiveInterval
public void setMaxInactiveInterval(int interval)

Specifies the time, in seconds, between client requests before the
servlet container will invalidate this session. A negative time indicates
the session should never timeout.
Parameters:
interval - An integer specifying the number of seconds

Re: Server-Side Includes (was Re: JSP and embedded servlet call)

[Kurt Bernhard Pruenner]

"Craig R. McClanahan" wrote:
> Kurt Bernhard Pruenner wrote:
> > No no no no no... I _don't_ want to add SHTML/SSI support to Tomcat - I 
> > want mod_jserv fixed so that, using the SSI facility in Apache,
> 
> Anyone who wishes to make this change (instead of or in addition to adding 
> SHTML support to Tomcat) is welcome to propose such a patch.  In open source 
> parlance, this is not an itch of mine that needs to be scratched ... :-)

Well, okay... looks like I'll have to rake through the code myself... just
thought somebody who already knows a bit more about mod_jserv would be
interested... *g*

But actually, I guess trying the same with mod_jk and probably patching that
instead (if it shows the same behaviour) would make even more sense - after
all, mod_jk is slated to replace mod_jserv in the future, isn't it?

-- 
Kurt Bernhard Pruenner --- Haendelstrasse 17 --- 4020 Linz --- Austria
Music: http://www.mp3.com/Leak --- Work: http://www.ssw.uni-linz.ac.at
...It might be written "Mindfuck", but it's spelt "L-A-I-N"...
np: LFO - Tied Up: Spiritualised Mix (WARP10+3.1 Remixes)

iis5 plugin and tc3.2.b7

[Milton]

Hi there!
I have been testing the TC3.2.b7 and the accompanied isapi_redirect.dll with w2k
server (sp1).  TC has been running real smooth as a standalone.  With iis5 and
the plugin, static stuff is real fast.  Great job on new virutal host using host
names or ips.  Have not run into any problems so far.

Here is the only problem I have with the plug in:
A response.sendRedirect() in the jsp code takes about 1 minute to get to
iis5.  Here is the scenario:
1. iis5log: GET /cart.jsp to isapi.dll - This jsp performs an update then does a
response.sendRedirect() to itself with diff parameters.
 after 1 minute 
2. iis5log: GET /cart.jsp to isapi.dll - This part handled the sendRedirect.
Instant reponse.

I reviewed the isapi detailed trace log and no errors.  There were no timestamps
in the log so I don't know where it stalled out.  TC standalone processes these
2 gets in 1 second which is normal

The cart.jsp code looks like this.
Begin of cart.jsp
debug.print(in cart time)
if param.action == buy then
   update the cart
   response.sendRedirect("cart.jsp?action=display")
   debug.print(redirect time)
   return
endif


End of cart.jsp


I put debug statements to trace the action at jsp entry and within the buy
logic.  It took about 1 minute after the response.sendRedirect command was exec
and the next entry into this jsp.  This corresponds to the web server log.
Please check it out and email me if you need me to provide more info or
whatever.

Thanks for the great job you folks are doing.

Thanks
Milton - [EMAIL PROTECTED]
Data Services of Hawaii, Inc.
1312 Kaumualii Street #A
Honolulu HI  96817
Phone (808) 841.7253

Re: Session timeout

[Craig R. McClanahan]

Charles Sabourdin wrote:

> --- "Craig R. McClanahan"
> <[EMAIL PROTECTED]> wrote:
> > "Ivan E. Markovic" wrote:
> >
> > > Quick question. How long are sessions inside
> > Apache/Tomcat valid for?
> > > In all the Servlet documentation I have it is
> > stated that session
> > > duration is server specific. So I checked the
> > Tomcat and Apache docs
> > > and could find no indicator as to what that
> > duration was. Anyone able
> > > to enlighten me?
> > >
> >
> > You can set the default timeout in your web.xml
> > file, using the 
> > element (see the spec, or one of the books or
> > articles that talk about the
> > web.xml format).
> >
> > If you do not specify it, Tomcat defaults to 30
> > minutes.
> >
> > Note that you can also change the timeout on a
> > session dynamically, by calling
> > the session.setMaxInactiveInterval() method.
> >
> > > I v a n ...
> >
> > Craig McClanahan
>  but in that  tag what is the unit ?
> secondes, minutes?
>

Download the spec, and you will see that it's defined in minutes.  You are going
to need the spec anyway to get the syntax and placement of this element correct.

http://java.sun.com/products/servlet/download.html

Craig

Re: System properties

[Craig R. McClanahan]

Miles Daffin wrote:

> > >
> > > As long as this property value is never broadcast
> > > i.e. System.setProperty("myProperty", "42");
> > >
> >
> > System properties, in the sense that we are talking about the
> java.lang.System
> > class, are global to the entire JVM.  Therefore, even if the statement
> above was
> > allowed, it would set the property "myProperty" for *all* web applications
> that
> > are installed in a single Tomcat instance, and would therefore cause a
> name
> > clash if another app also tried to set the value for "myProperty".
>
> Are such System calls prevented (intercepted somehow) in TC?
>

If you are running under JDK 1.1, they are prevented by the fact that JDK 1.1
does not have this method.

If you are running Tomcat 3.2 under a security manager, you can configure your
security policies to prevent this if you want to -- otherwise there is no
restriction.

>
> > Context parameters, on the other hand, are unique to an individual web
> > application so there are no problems with two web-apps having parameters
> (or
> > anything else, for that matter) with the same names.
>
> I shall use them from now on.
>
> 
>
> > You can use static variables for this purpose and get away with it, as
> long as
> > the classes involved are loaded from WEB-INF/classes or WEB-INF/lib.  The
> reason
> > for this is that each web application is loaded by its own class loader,
> so
> > statics created by such classes are visible only within the web app. I
> have a
> > personal bias against statics (I prefer to store global application
> objects in
> > the servlet context, so that they are easily visible to all servlets and
> JSP
> > pages in my apps).
>
> Do you mean that you have a class, say Contants, that has a bunch of statics
> (public contants and maybe methods) that is only visible to objects in a
> particular
> app? In which case you are still using statics - only the scope has been
> narrowed,
> for obviously good reasons.
>

Let's say I have class com.mycompany.mypackage.MyClass that has a static method
foo().  Let's say that I install this class under WEB-INF/classes in two
different web apps.

In this scenario, there are two instances of MyClass in memory (even though they
have the same name), because they are loaded by two different class loaders.

The only way to have a static global across web apps in Tomcat is to put it on
the system classpath -- but beware:  the ability to do this portably is *not*
guaranteed by the servlet spec.

>
> >
> > > Miles
> >
> > Craig
> >
> Miles

Craig

Re: Server-Side Includes (was Re: JSP and embedded servlet call)

[Craig R. McClanahan]

Kurt Bernhard Pruenner wrote:

>
> No no no no no... I _don't_ want to add SHTML/SSI support to Tomcat - I want
> mod_jserv fixed so that, using the SSI facility in Apache,

Anyone who wishes to make this change (instead of or in addition to adding SHTML
support to Tomcat) is welcome to propose such a patch.  In open source parlance,
this is not an itch of mine that needs to be scratched ... :-)

Craig

Tomcat StartUp Problem

[Kumar Shailendra]

Hi,
Can somebody please help me solve the error I am getting  I try to start

Tomcat on my machine.
"Java.lang.ClassNot  FoundException:
org/apache/tomcat/service/http/HttpConnectionHandler" .
In short I get Class not found exception for HttpConnectionHandler.
I have set the TOMCAT_HOME and JAVA_HOME variable.My OS is Windows NT
workstation 4.2 .

Thanks & Regards,
Shailendra

Re: Session timeout

[Charles Sabourdin]

--- "Craig R. McClanahan"
<[EMAIL PROTECTED]> wrote:
> "Ivan E. Markovic" wrote:
> 
> > Quick question. How long are sessions inside
> Apache/Tomcat valid for?
> > In all the Servlet documentation I have it is
> stated that session
> > duration is server specific. So I checked the
> Tomcat and Apache docs
> > and could find no indicator as to what that
> duration was. Anyone able
> > to enlighten me?
> >
> 
> You can set the default timeout in your web.xml
> file, using the 
> element (see the spec, or one of the books or
> articles that talk about the
> web.xml format).
> 
> If you do not specify it, Tomcat defaults to 30
> minutes.
> 
> Note that you can also change the timeout on a
> session dynamically, by calling
> the session.setMaxInactiveInterval() method.
> 
> > I v a n ...
> 
> Craig McClanahan
 but in that  tag what is the unit ?
secondes, minutes?

__
Do You Yahoo!?
Yahoo! Calendar - Get organized for the holidays!
http://calendar.yahoo.com/

Re: System properties

[Miles Daffin]

> >
> > As long as this property value is never broadcast
> > i.e. System.setProperty("myProperty", "42");
> >
>
> System properties, in the sense that we are talking about the
java.lang.System
> class, are global to the entire JVM.  Therefore, even if the statement
above was
> allowed, it would set the property "myProperty" for *all* web applications
that
> are installed in a single Tomcat instance, and would therefore cause a
name
> clash if another app also tried to set the value for "myProperty".


Are such System calls prevented (intercepted somehow) in TC?

> Context parameters, on the other hand, are unique to an individual web
> application so there are no problems with two web-apps having parameters
(or
> anything else, for that matter) with the same names.

I shall use them from now on.



> You can use static variables for this purpose and get away with it, as
long as
> the classes involved are loaded from WEB-INF/classes or WEB-INF/lib.  The
reason
> for this is that each web application is loaded by its own class loader,
so
> statics created by such classes are visible only within the web app. I
have a
> personal bias against statics (I prefer to store global application
objects in
> the servlet context, so that they are easily visible to all servlets and
JSP
> pages in my apps).

Do you mean that you have a class, say Contants, that has a bunch of statics
(public contants and maybe methods) that is only visible to objects in a
particular
app? In which case you are still using statics - only the scope has been
narrowed,
for obviously good reasons.

>
> > Miles
>
> Craig
>
Miles

message in mod_jk log file.

[Tony Mueller]

Hello,

My webapp seems to run fine on apache - mod_jk - tomcat 3.1 except that I 
occasionally get the following message in my mod_jk log file?

[jk_ajp12_worker.c (596)]: ajpv12_handle_response, error writing back to
server

Has anyone else got the same message and knows how to resolve it?

Thanks in advance

Tony Mueller

Would you please post your config files

[Debbie Cao]

Hi, Larry,
I am configuring Apache, Tomcat on NT 4.0 Workstation. The configuration
is really a pain. Your message helped a lot. There are still a few questions

not very clear to me:
1. The port numbers. Where should I say 8080, where to say 80, 8007,
8040, localhost, ...
2. How didyou configure Tomcat using Ajpv12 protocol?
3. JServ has the concept of "zone". Does Tomcat support this feature?
So, I am very much interested in reading your config files. Would you please
post your configu files. Or, if you don't mind, please send them to:
[EMAIL PROTECTED]
or  [EMAIL PROTECTED]
Thanks a lot,
Dehe Cao

Questions re: Tomcat and IBM VAJ

[Lhohmann]

Hello. I'm trying to get Tomcat running under Windows 'NT with IBM VAJ. 
Unfortunately, I haven't been able to make everything work together just yet, 
and would greatly appreciate your help in resolving what is probably a very 
simple problem.

Environment
 - MS-Windows 'NT 4.0
 - IBM VAJ version 3.0 for J2SE 1.2 (the free, up to 750 classes version of 
VAJ)
 - Tomcat 3.1

All of the above have been downloaded and installed into their "default" 
locations. Environment variables include the following:
CLASSPATH=C:\IBMVJava2\eab\runtime30
JAVA_HOME=C:\IBMVJava2\eab\runtime30
TOMCAT_HOME=C:\jakarta\tomcat

When I enter the following commands:
C:\jakarta\tomcat\bin>startup
Starting tomcat in new window
Using classpath: 
c:\jakarta\tomcat\classes;c:\jakarta\tomcat\lib\webserver.jar;c
:\jakarta\tomcat\lib\jasper.jar;c:\jakarta\tomcat\lib\xml.jar;c:\jakarta\tomca

t\
lib\servlet.jar;c:\j2sdkee1.2.1\lib\tools.jar;C:\IBMVJava2\eab\runtime30;
C:\jakarta\tomcat\bin>

A new MS-DOS Command Prompt window opens with the title 
"c:\WINNT\system32\java.exe". It contains the following output, presumably 
from Tomcat:
Context log: path="/examples" Adding context path="/examples"  
docBase="webapps/
examples"
Context log: path="" Adding context path=""  docBase="webapps/ROOT"
Context log: path="/test" Adding context path="/test"  docBase="webapps/test"
Starting tomcat. Check logs/tomcat.log for error messages
Starting tomcat install="c:\jakarta\tomcat" home="C:\jakarta\tomcat" 
classPath="
c:\jakarta\tomcat\classes;c:\jakarta\tomcat\lib\webserver.jar;c:\jakarta\tomca

t\
lib\jasper.jar;c:\jakarta\tomcat\lib\xml.jar;c:\jakarta\tomcat\lib\servlet.jar

;c
:\j2sdkee1.2.1\lib\tools.jar;C:\IBMVJava2\eab\runtime30;"
Context log: path="/admin" Automatic context load 
docBase="C:\jakarta\tomcat\web
apps\admin"
Context log: path="/admin" Adding context path="/admin"  
docBase="C:\jakarta\tom
cat\webapps\admin"

The logs do not seem to indicate an error message -- the tomcat.log contains 
the following:
Starting endpoint port="8080" 
handler="org.apache.tomcat.service.http.HttpConnectionHandler" Starting 
endpoint port="8007" 
handler="org.apache.tomcat.service.connector.Ajp12ConnectionHandler" 


While this seems to indicate that Tomcat is working properly, I know it 
isn't. I have two files, simple.html and simple.jsp (downloaded from Wrox, as 
I'm using their Professional JSP book as a reference). 
begin simple.html 


Simple JSP Example




How many times?








 end simple.html 
 begin simple.jsp 
<%@ page language="java" %>



Simple JSP Example





  <% int numTimes = Integer.parseInt(request.getParameter("numtimes"));
 for (int i = 0; i < numTimes; i++) { 
  %>
   Hello, world!
  <% }
  %>


<%@ include file="PageFooter.html" %>



 end simple.jsp 

simple.html loads fine. When I enter a number into the resultant form 
simple.jsp is loaded but the JSP servlet code IS NOT executed. What am I 
doing wrong?

Luke Hohmann

Re: Server-Side Includes (was Re: JSP and embedded servlet call)

[Kurt Bernhard Pruenner]

"Craig R. McClanahan" wrote:
> Kurt Bernhard Pruenner wrote:
> > "Lacerda, Wellington (AFIS)" wrote:
> > > Why don't you use the plain  ?
> >
> > Because the page is an shtml page that's handled by Apache, and since it's
> > possible to include JSP-pages that are handled by mod_jserv with the old 
> > JServ I'm curious why this shouldn't be possible with Tomcat. Also, all 
> > our html pages are actually treated as shtml pages, but we really don't 
> > want to
> >
> > a) change all of them to JSPs
> > b) have all of them handled by Tomcat for performance reasons
> >
> > (FYI, shtml pages are handled by mod_rewrite for Apache, and #include
> > virtual="..." allows you to include any other page that can otherwise be
> > served by Apache, except that it mysteriously doesn't work with pages that
> > come from Tomcat, while it _does_ work with the older JServ... that's why
> > we've got both JServ and Tomcat running on our server currently, which of
> > course isn't really preferable... :/)
> 
> It would be feasible to add SHTML support to Tomcat.

No no no no no... I _don't_ want to add SHTML/SSI support to Tomcat - I want
mod_jserv fixed so that, using the SSI facility in Apache, I can include
Tomcat-generated pages that get passed back to Apache via mod_jserv - the only
thing that hinders this is that mod_jserv seems to pass the wrong location to
Tomcat when it's called to include a page, as it passes the location of the
including (SHTML) page instead of the (JSP's) location that's given using the
"virtual" attribute in the #include statement of the SHTML page... you can see
that when Tomcat generates a file not found message, which is correctly
included into the SHTML page.

I, for performance reasons, don't want to serve all HTML pages which might or
might not include SSI statements using Tomcat, I just want mod_jserv for
Tomcat to work with Apache's SSI parser like mod_jserv for JServ did. :(

Am I really describing it so bad that nobody seems to understand me? :(

I'll try describing it in detail:

1. A request for an SHTML file goes to Apache
2. Apache's SSI facility starts parsing the SHTML file
3. Said facility encounters an 

Re: System properties

[Craig R. McClanahan]

Miles Daffin wrote:

> Hi guys,
>
> > > From the looks of things, it is just the way it is, although I was
> hoping
> > > the there would be some way to put stuff in the web.xml file, maybe in
> the
> > > servlet tag like:
> > >
> > > 
> > > myproperty
> > > 42
> > > 
> > >
> > > Anyone think this is a good idea, or am I talking rubbish?
> >
> > It's a bad idea.  Keep your web-apps encapsulated.
>
> Why would such a move, if it were possible, break encapsulation?
>
> As long as this property value is never broadcast
> i.e. System.setProperty("myProperty", "42");
>

System properties, in the sense that we are talking about the java.lang.System
class, are global to the entire JVM.  Therefore, even if the statement above was
allowed, it would set the property "myProperty" for *all* web applications that
are installed in a single Tomcat instance, and would therefore cause a name
clash if another app also tried to set the value for "myProperty".

Context parameters, on the other hand, are unique to an individual web
application so there are no problems with two web-apps having parameters (or
anything else, for that matter) with the same names.

>
> If you want application local properties then simply create a class to hold
> them
> with a static method for obtaining them from anywhere in your app (like
> System.getProperties()).
>

You can use static variables for this purpose and get away with it, as long as
the classes involved are loaded from WEB-INF/classes or WEB-INF/lib.  The reason
for this is that each web application is loaded by its own class loader, so
statics created by such classes are visible only within the web app.  I have a
personal bias against statics (I prefer to store global application objects in
the servlet context, so that they are easily visible to all servlets and JSP
pages in my apps).

> Miles

Craig

Problems Installing Apache,tomcat,cocoon and php4

[Stamaths Karvounarakhs]

I have tried to install apache1.3.12, tomcat3.1, cocoon1.7 and php4.0.1pl2
in a Solaris2.8.Those who have tried to install cocoon in Solaris must know that it's 
not
easy at all, but at last guided by the install.case.solaris in the docs/
directory of cocoon1.8 I made it.What I can't make work is php with apache
and cocoon.(When I do not include the tomcat-apache-cocoon.conf in the
httpd.conf, php works fine but when I include it apache doesn't seem to 
understand what to do with the .php file).PLEASE HELP ME!!

~user request Error: 404 Location: /~bryan File Not Found /~bryan

[bryan]

How do I get a standalone Tomcat 3.1 server to resolve ~user requests?  The
Apache 1.3.11 server handles ~user requests fine (UserDir public_html).  Is
there something that needs to be added to jakarta-tomcat/conf/server.xml or
a command line argument that lets Tomcat know what '~' expand to?  '~'
expands to /dc/home NOT /home on our systems.

Example:  http://asterix:8081/~bryan

Apache handles this fine.  Tomcat complains that ~bryan directory is not found.

Thanks.

Bryan
-- 

Re: Server-Side Includes (was Re: JSP and embedded servlet call)

[Craig R. McClanahan]

Kurt Bernhard Pruenner wrote:

> "Lacerda, Wellington (AFIS)" wrote:
> > Why don't you use the plain  ?
>
> Because the page is an shtml page that's handled by Apache, and since it's
> possible to include JSP-pages that are handled by mod_jserv with the old JServ
> I'm curious why this shouldn't be possible with Tomcat. Also, all our html
> pages are actually treated as shtml pages, but we really don't want to
>
> a) change all of them to JSPs
> b) have all of them handled by Tomcat for performance reasons
>
> (FYI, shtml pages are handled by mod_rewrite for Apache, and #include
> virtual="..." allows you to include any other page that can otherwise be
> served by Apache, except that it mysteriously doesn't work with pages that
> come from Tomcat, while it _does_ work with the older JServ... that's why
> we've got both JServ and Tomcat running on our server currently, which of
> course isn't really preferable... :/)
>

It would be feasible to add SHTML support to Tomcat.

What this would take is creating a servlet to do the appropriate parsing and
processing, and map it to the "*.shtml" extension.  You might even be able to
start with the Apache JSSI servlet (at ) and modify it as
necessary to reflect the changes in the servlet API since 2.0.

However, no one to date has volunteered to add this feature.  If someone wants to,
I'd be happy to integrate it in.  Any takers?

>
> --
> Kurt Bernhard Pruenner --- Haendelstrasse 17 --- 4020 Linz --- Austria
>

Craig McClanahan

Re: Servlet context

[Craig R. McClanahan]

David Oxley wrote:

> We have written a servlet that is being used in a virtual host setup. I need
> to retrieve the virtual host that the servlet has been setup up to work
> against. I can retrieve the server name but not the virtual host name. Also
> I would like to retrieve the context.
> We are using Apache 1.3.14 mod_jk'ed with Tomcat 3.2b7 using ajp13.
>

I recently committed a fix for this -- request.getServerName() will return the
virtual host name, as of 3.2b8 (to be created tonight, if all goes well).

>
> Dave
> [EMAIL PROTECTED]

Craig McClanahan

Re: Session timeout

[Craig R. McClanahan]

"Ivan E. Markovic" wrote:

> Quick question. How long are sessions inside Apache/Tomcat valid for?
> In all the Servlet documentation I have it is stated that session
> duration is server specific. So I checked the Tomcat and Apache docs
> and could find no indicator as to what that duration was. Anyone able
> to enlighten me?
>

You can set the default timeout in your web.xml file, using the 
element (see the spec, or one of the books or articles that talk about the
web.xml format).

If you do not specify it, Tomcat defaults to 30 minutes.

Note that you can also change the timeout on a session dynamically, by calling
the session.setMaxInactiveInterval() method.

> I v a n ...

Craig McClanahan

Re: Serving welcome page v. issuing HTTP redirect

[Phil Radden]

On Mon, 20 Nov 2000, Craig R. McClanahan wrote:
> The redirect should be setting an absolute URL in the "Location:" header -- if
> it's not, that is a Tomcat bug, because the container is required to create an
> absolute URL even if the program specifies a relative one.  I will look into it.

Many thanks Craig - and there's no need to investigate, because it appears I
was confused.  Tomcat 3.2 does the right thing (in my testing at least) -
the system that wasn't was using another platform which had better remain
anonymous!

Apologies, and thanks for the swift explanation

Re: Future deveploment of Tomcat

[Craig R. McClanahan]

Christian Mallwitz wrote:

> Hi,
>
> Knowing that Tomcat is an onging public effort and that most the people
> working on it are volunteers I would like know if there is something like a
> "release schedule". I've seen discussion about Tomcat 3.3 but  3.2 is still
> beta (working great although) and 4.0 is available as milestone builds.
>
> Will 3.2 release before Christmas?

I am going to create "beta 8" of 3.2 tonight (unless some unforseen new bug pops
up).  If there are no critical bugs found in it (probably a week of testing will
be sufficient), we will release 3.2 final shortly after that.  At worst, I would
expect to see no more than one more beta round.

NOTE:  There will be some outstanding issues with the MOD_JK and AJP13 web
connectors that will not have been fixed in 3.2 final, because no one with the
appropriate knowledge has volunteered to find the problems and propose patches
for them.

>
> What is the plan about 4.0 (Q1 or Q2 2001)?
>

Timing for 4.0 is less certain than for 3.2.

Current thinking is to release a "milestone 5" within the next couple of weeks,
as soon as the web connector code has been checked in and at least minimally
tested.  After that, I am going to suggest a feature freeze on 4.0 and start
doing beta cycles to fix remaining bugs (plus a few low-risk performance
improvements).

Personally, I would love to havea release-quality Tomcat 4.0 for Christmas.
Reality is more likely to be early in 2001.

NOTE:  The servlet 2.3 and JSP 1.2 specifications are *not* final at this point,
and most likely will not be by the time 4.0 is released.  Therefore, we might
need to do some point releases (4.0.1, etc.) to adapt to any changes in the
specs that require Tomcat to be updated.

As soon as 4.0 goes into beta cycles, we will also need to start planning on
what we want to do for 4.1, and when.  These discussions will take place on the
TOMCAT-DEV mailing list, so feel free to subscribe to that list to participate
in the discussions, and/or to stay informed.

>
> Thanks
> Christian

Craig McClanahan

Re: Serving welcome page v. issuing HTTP redirect

[Craig R. McClanahan]

Phil Radden wrote:

> There appears to be a change of behaviour between Tomcat 3.1 and 3.2bX
> regarding the handling of the top page for a servlet.
>
> I have a servlet in ROOT context, and wish a particular file, say index.jsp,
> to be used for requests for 'GET / HTTP/1.0'.  In 3.1, this is what happens.
> In 3.2, an HTTP redirect to 'Location: /index.jsp' is issued instead.
>
> Ideally, I'd want the 3.1 behaviour.  And failing that, the redirection
> should really be to a fully-qualified URL.  Any suggestions on what I should
> try?
>

What's going on is Tomcat is redirecting to the welcome page you have configured
for this directory.  The redirect is important because a welcome file can have
relative hyperlinks, and they are resolved to the wrong place under several
different usage cases.

The redirect should be setting an absolute URL in the "Location:" header -- if
it's not, that is a Tomcat bug, because the container is required to create an
absolute URL even if the program specifies a relative one.  I will look into it.

> Many thanks

Craig McClanahan

Opinions on Websphere?

[Daniel A. Theobald]

I posted this on the webspheres newsgroup, but only got one opinion..
wondering if there are more opinions out there

I have some general questions I am hoping I can get some opinions
on...see the bottom of my message for info on making an easy $4000...

Background: 
Vecna Technologies was chosen by DOD to help define the World-wide
E-Health strategy and prototype the first phase of the project.  We are
well underway and have already completed several components such as
online appointments, online enrollment, desktop customization, web
wizards for content creation and management (there will be well over 1
million distinct content pages generated by providers, clinics,
hospitals, regions, etc.)

The site will also incorporate health content from a big vendor such as
Healtheon...

Questions:

1. How does websphere compare to other portal creation tools?  Are there
others you would recommend looking at?

2. Many of our components currently are written using java 1.3.   I have
installed the trial version of websphere for linux and it seems to need
java 1.8.8.  Anyone know if/when it will support the java 2 platform?

3. Anyone know of sites that have been built using websphere that I
might refer to?

4. How does it maintain state? (Can it work with cookies disabled?)

5. Is it dependent on using Frames in any way?


This will be one of the largest healthcare portals in the world, so it
is crucial that we choose the right application server platform

thanks,
daniel

PS. Vecna is looking for top notch java programmers and high level
architects. Vecna has the best salaries and benefits to be found, huge
profit sharing AND equity, and a unique company culture emphasizing
community service (such as open-source).  Vecna is located in the DC
area (No VA, DC, Baltimore) and in Boston, but many of our team members
work remotely, so location is not an issue  see http://www.vecna.com

$4000.00 referral fee!  

Anyone that refers someone to us that we hire will be paid a $4000
referral fee, so let us know if you know someone that may be a good fit.

We also use consultants regularly, so send your resume even if you are
just looking for short-term or part-time

Re: Servlet invoker woes - servlet-mapping /servlets/* to invoker fails on 3.2

[Craig R. McClanahan]

Rachel Greenham wrote:

> On Saturday 18 November 2000 00:29, you wrote:
> > As of Tomcat 3.2-b8 (to be created on Monday), you will be able to set a
> > parameter in the server.xml file to override the prefix used by the invoker
> > servlet.  There is an entry that looks like this:
> >
> >  > className="org.apache.tomcat.request.InvokerInterceptor"
> > debug="0" prefix="/servlet/" />
> >
> > and you can change the prefix attribute to "/servlets/" for your particular
> > case.
>
> I already did this, on 3.2b7, the same solution works already. That's how I
> resolved the issue for now, but I wasn't happy with it, because our webapps
> weren't compatible between 3.1 and 3.2, which was worrying to say the least.
> What's needed is a *standard* solution to this problem that will work across
> all API-compliant servlet containers - ie: defined in the WEB-INF/web.xml for
> the webapp. Or I need to know for sure that there is no such standard
> solution and we should be using aliases for all our servlets. I had another
> look through the API specification and to me it looks like the latter, which
> would be a major pain.
>

As I have written several times before, there is *no* standard for the invoker
service (/servlet/* or /servlets/*).  It is *not* part of the servlet
specification -- if you'd like to encourage it to be added, you should send mail
to the address on the spec ([EMAIL PROTECTED]).

What *is* in the standard is the servlet mappings that are defined in the 2.2
specification, which you can download at
.  This is the only thing you
can count on being portable.

>
> > In Tomcat 4.0, the ability to configure global defaults in the
> > "conf/web.xml" file is restored to the way it worked in Tomcat 3.1 -- it is
> > much more intuitive for things like this.
>
> Because we're trying to write nice well-behaved webapps, we don't want to
> have to change a global setting, only a local-to-webapp setting. I was able
> to do this on Tomcat 3.1.
>

Tomcat 3.2 was changed (not by me!) in ways that were gratuitously incompatible
with 3.1.  It's too late in the release cycle for 3.2 to do much about this, but
you'll be happy to know that Tomcat 4.0 returns to the approach used in Tomcat
3.1 (using the "conf/web.xml" file as global defaults for all web apps).

Note, however, that reliance on this configuration mechanism is Tomcat specific.
Basically, you should look at things this way:  "things inside my web app's
/WEB-INF/web.xml file are portable -- anything else is not.

> Rachel

Craig McClanahan

UnsupportedEncodingException Starting Tomcat

[barreiro]

Hi,

I've problems starting Tomcat in a Linux Box with:

RedHat 6.2
apache-1.3.12
JDK118_v3
jakarta-tomcat-3.2-b7

startup.sh script dead with the following messages:
-
java.io.UnsupportedEncodingException: UTF8
at java.lang.Throwable.(Throwable.java:40)
at java.lang.Exception.(Exception.java:21)
at java.io.IOException.(IOException.java:22)
at
java.io.UnsupportedEncodingException.(UnsupportedEncodingException.java:24
)
at
kaffe.io.ByteToCharConverter.getConverter(ByteToCharConverter.java:89)
at java.io.InputStreamReader.(InputStreamReader.java:37)
at com.sun.xml.parser.XmlReader.useEncodingDecl(XmlReader.java:316)
at com.sun.xml.parser.XmlReader.(XmlReader.java:248)
at com.sun.xml.parser.XmlReader.createReader(XmlReader.java:86)
at com.sun.xml.parser.InputEntity.init(InputEntity.java:139)
at com.sun.xml.parser.Parser.parseInternal(Parser.java:463)
at com.sun.xml.parser.Parser.parse(Parser.java:284)
at javax.xml.parsers.SAXParser.parse(SAXParser.java:155)
at javax.xml.parsers.SAXParser.parse(SAXParser.java:126)
at org.apache.tomcat.util.xml.XmlMapper.readXml(XmlMapper.java:214)
at org.apache.tomcat.startup.Tomcat.execute(Tomcat.java:189)
at org.apache.tomcat.startup.Tomcat.main(Tomcat.java:237)
FATAL: configuration error
java.lang.Exception: Can't open config file: ./../conf/server.xml due to:
java.io.UnsupportedEncodingException: UTF8
at java.lang.Throwable.(Throwable.java:40)
at java.lang.Exception.(Exception.java:21)
at org.apache.tomcat.util.xml.XmlMapper.readXml(XmlMapper.java:223)
at org.apache.tomcat.startup.Tomcat.execute(Tomcat.java:189)
at org.apache.tomcat.startup.Tomcat.main(Tomcat.java:237)
---

A similar configuration work fine in my developer box.

Anyone can help me? 

Regards,

p.d. LANG is set to es_ES

Manuel de Vega Barreiro. [EMAIL PROTECTED]   En un lugar de la red
Madrid. España.de cuyo nombre no quiero
http://www.croftj.net/~barreiro Linux Landia   acordarme..

UnsupportedEncodingException Starting Tomcat

[barreiro]

Hi,

I've problems starting Tomcat in a Linux Box with:

RedHat 6.2
apache-1.3.12
JDK118_v3
jakarta-tomcat-3.2-b7

startup.sh script dead with the following messages:
-
java.io.UnsupportedEncodingException: UTF8
at java.lang.Throwable.(Throwable.java:40)
at java.lang.Exception.(Exception.java:21)
at java.io.IOException.(IOException.java:22)
at
java.io.UnsupportedEncodingException.(UnsupportedEncodingException.java:24
)
at
kaffe.io.ByteToCharConverter.getConverter(ByteToCharConverter.java:89)
at java.io.InputStreamReader.(InputStreamReader.java:37)
at com.sun.xml.parser.XmlReader.useEncodingDecl(XmlReader.java:316)
at com.sun.xml.parser.XmlReader.(XmlReader.java:248)
at com.sun.xml.parser.XmlReader.createReader(XmlReader.java:86)
at com.sun.xml.parser.InputEntity.init(InputEntity.java:139)
at com.sun.xml.parser.Parser.parseInternal(Parser.java:463)
at com.sun.xml.parser.Parser.parse(Parser.java:284)
at javax.xml.parsers.SAXParser.parse(SAXParser.java:155)
at javax.xml.parsers.SAXParser.parse(SAXParser.java:126)
at org.apache.tomcat.util.xml.XmlMapper.readXml(XmlMapper.java:214)
at org.apache.tomcat.startup.Tomcat.execute(Tomcat.java:189)
at org.apache.tomcat.startup.Tomcat.main(Tomcat.java:237)
FATAL: configuration error
java.lang.Exception: Can't open config file: ./../conf/server.xml due to:
java.io.UnsupportedEncodingException: UTF8
at java.lang.Throwable.(Throwable.java:40)
at java.lang.Exception.(Exception.java:21)
at org.apache.tomcat.util.xml.XmlMapper.readXml(XmlMapper.java:223)
at org.apache.tomcat.startup.Tomcat.execute(Tomcat.java:189)
at org.apache.tomcat.startup.Tomcat.main(Tomcat.java:237)
---

A similar configuration work fine in my developer box.

Anyone can help me? 

Regards,

p.d. LANG is set to es_ES

Manuel de Vega Barreiro. [EMAIL PROTECTED]   En un lugar de la red
Madrid. España.de cuyo nombre no quiero
http://www.croftj.net/~barreiro Linux Landia   acordarme..

RE: One user -> 2 parallel threads

[Steve Toth]

You could also make the jsp singlethreaded.

<%@ page isSingleThreaded="false" %>

This will tell tomcat that the jsp should only be in one thread, multiple
requests will be queued up. thus the jsp should be FAST (no IO/DB access).
Same as implementing the servlet Interface SingltThreadModel.

 -Original Message-
From: Till Gartner [mailto:[EMAIL PROTECTED]]
Sent: Saturday, November 18, 2000 7:32 AM
To: '[EMAIL PROTECTED]'
Subject: AW: One user -> 2 parallel threads


You're right: We checked the Apache log and the user just double clicked.

But Netscape seems to have a bug that cannot be handled with the JavaScript
solution:
If you have your security settings in a way that he comes up with a
messagebox before sending out POSTs, he sometimes "overhears" the first
click in that box and only reacts on the second click. Strangely he then
also sends out 2 requests. 

How does the JavaScript to catch this kind of doble clicks look like?

Thanks for your help ona rainy saturday,
-- Till.

-Ursprüngliche Nachricht-
Von: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED]]
Gesendet: Samstag, 18. November 2000 16:21
An: [EMAIL PROTECTED]
Betreff: Re: One user -> 2 parallel threads


So, in those cases, I've built JavaScript logic running in the browser (in 
the HTML file) that will set a boolean flag to block the second (and 
subsequent) click(s).  Is there a simple way to handle this server-side (I 
can imagine the hard ways!), or does anyone know that the client side still 
has holes I have not witnessed?

JT

<< Undoubtedly the user clicked twice on the button or link and
 actually created two requests.>>

 > 
 > Hi mail list,
 > 
 > we encounter the following problem with Tomcat. From time to time a
browser
 > request seems to be handled by two threads from tomcat in parallel. Our
own
 > log from java servlets shows entries of different threads doing the same
 > task for the same user, wich are just some milli secs apart. >>

problem with applet and JDBC

[Harding, David]

tomcat 3.1, mysql 3.23.21, mm.mysql 2.02, freebsd 4.0-RELEASE

we have an applet that tries to connect our database (applet and database
hosted on same machine) using a connection pooling class (a la Core Servlets
and JavaServer Pages by Marty Hall)problem is that we get a "Connection
failure during handshake" exception message (thrown by mm.mysql) and
*sometimes* a corresponding message in mysql error log that a connection was
aborted due to an error reading communication packets.  a servlet that runs
on the same machine as the database and hosts the applet and uses the same
connection pooling class never fails.  *ag!*

this is an ongoing problem and we are still working on determining if our
network is the problem (very unlikely from what we have seen so far, but
still looking into it).

mysql documentation has a section about this error and states that it could
be due to a couple of things:
* inactivity on the connection (we have not even connected yet)
* client exited without calling close() (well this may be where the
error in the mysql log comes from, but is not the source of the problem)
* bad TCP/IP configuration (problem can be reproduced on any machine)
* error on read in thread library (well, maybe, but never had a problem
before and ran mysql in "--one-thread" mode and still got the error)
* router, hub, switch, cable, NIC problem.hard to diagnose and we
have had no other network problems (network is *very* well used)

i am stumped.  any ideas at all are greatly appreciated.

cheers,
   david j harding

Re: Anyway to render JSP to static HTML? -- What happens with Cocoon XSP servlet examples?

[Derek DeMoro]

Just set up a URL to point to the url where the jsp lies and 
save the results.
 
ex. 
 
Just write out the string returned to a file.

/** *  HttpPing.  Used to ping specified urls. * 
 * @author Derek DeMoro */package jsp.beans;
import java.net.URL;import java.io.InputStreamReader;import 
java.io.BufferedReader;import java.net.MalformedURLException;
public class HttpPing extends Object{
 /** * Get html from specified url. * @param url - The 
url to ping. */  public static String getUrlResults(String 
url)  {    String urlString = 
null;    StringBuffer urlBuffer = new StringBuffer();
    try    
{  //Hand the Url Object a URL 
address.  URL theUrl = new URL(url);
  try  
{    //Open an input stream from the 
URL    BufferedReader in = new 
BufferedReader(new InputStreamReader(theUrl.openStream()));
    
try    
{  //Read Contents From 
URL  while ((urlString 
= in.readLine()) != 
null)  
{    
urlBuffer.append(urlString);    
urlBuffer.append("\n");  
}    
}    catch(Exception 
ioEx)    
{  
ioEx.printStackTrace();    
}  }  
catch(Exception ex)  
{    
ex.printStackTrace();  }   
}   catch(MalformedURLException malformedEx)   
{  
malformedEx.printStackTrace();   }   return 
urlBuffer.toString();  }}
 
Derek DeMoroSenior Member Of Technical Staff For Oracle 
JdeveloperOracle Corp.
 

  - Original Message - 
  From: 
  cH3rN0b|L 
  To: [EMAIL PROTECTED] 
  
  Sent: Monday, November 20, 2000 9:21 
  AM
  Subject: Anyway to render JSP to static 
  HTML? -- What happens with Cocoon XSP servlet examples?
  
  Hi all!
  I was wondering if we can retake the topic I read 
  sometime ago in the list
  talking about saving JSP results to static 
  HTML.
  I'm hard behind that, but I have no luck with the 
  answers.
  I'm using Apache 1.3.12, Tomcat 3.2b5 and Cocoon 
  1.8 on RedHat 7.0,
  working toghether, 
  after some hard work to configure it.
  I need ,in some cases, JSP 
  results stored in static html files
  instead of sending the results to the client 
   internet navigator.
   
  Another question could be why in Tomcat 3.2b5 
  isn't included the
  xml.jar file, that suspectly starts XSP in 
  Cocoon, because I couldn't
  ever start that XSP servlet examples that comes with Cocoon.
  TIA for any answer that could help 
  me.
   
  Iván FontánTechnical Department
   
  LATINIA CORPORATION www.latinia.com
   
  Spain OfficeBarcelonaTrav. Gràcia, 
  342-34408025 Barcelona (Spain)Tel.: +34 934 465 026
   
  US OfficesNew York885 3rd AvenueSuite 
  2800New York, NY 10022
   
  Miami1210 Brickell AvenueSuite 
  430Miami, FL 33131

Re: Server-Side Includes (was Re: JSP and embedded servlet call)

[Kurt Bernhard Pruenner]

"Lacerda, Wellington (AFIS)" wrote:
> Why don't you use the plain  ?

Because the page is an shtml page that's handled by Apache, and since it's
possible to include JSP-pages that are handled by mod_jserv with the old JServ
I'm curious why this shouldn't be possible with Tomcat. Also, all our html
pages are actually treated as shtml pages, but we really don't want to 

a) change all of them to JSPs
b) have all of them handled by Tomcat for performance reasons

(FYI, shtml pages are handled by mod_rewrite for Apache, and #include
virtual="..." allows you to include any other page that can otherwise be
served by Apache, except that it mysteriously doesn't work with pages that
come from Tomcat, while it _does_ work with the older JServ... that's why
we've got both JServ and Tomcat running on our server currently, which of
course isn't really preferable... :/)

-- 
Kurt Bernhard Pruenner --- Haendelstrasse 17 --- 4020 Linz --- Austria
Music: http://www.mp3.com/Leak --- Work: http://www.ssw.uni-linz.ac.at
...It might be written "Mindfuck", but it's spelt "L-A-I-N"...
np: Avalon - Ladybug (ambient.01@hyperreal)

RE: Antwort: RE: tomcat vs orion vs weblogic testing results (Load Runner)

[David Oxley]

Oops. Tomcat twice the speed of WebSphere.

-Original Message-
From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED]]
Sent: 20 November 2000 15:42
To: [EMAIL PROTECTED]
Subject: Antwort: RE: tomcat vs orion vs weblogic testing results (Load
Runner)



Hi,
which one is faster, Tomcat or Websphere?

Thomas





David Oxley <[EMAIL PROTECTED]> am 20.11.2000 15:15:02

Bitte antworten an [EMAIL PROTECTED]

An:"'[EMAIL PROTECTED]'" <[EMAIL PROTECTED]>
Kopie:
Thema: RE: tomcat vs orion vs weblogic testing results (Load Runner)


Are there any benchmarks between Tomcat and WebSphere because at our site
we
are seeing Tomcat is twice the speed of Tomcat and I'm sure this can't be
right!!
Dave
[EMAIL PROTECTED]

-Original Message-
From: Amrhein, Thomas [mailto:[EMAIL PROTECTED]]
Sent: 20 November 2000 11:38
To: [EMAIL PROTECTED]
Subject: AW: tomcat vs orion vs weblogic testing results (Load Runner)


Which JVM did you use?


-Ursprüngliche Nachricht-
Von: Manne Anliot [mailto:[EMAIL PROTECTED]]
Gesendet am: Montag, 20. November 2000 11:15
An: '[EMAIL PROTECTED]'
Betreff: tomcat vs orion vs weblogic testing results (Load Runner)

Hi all..

I'm doing work load testing on a webapplication built on javascript, .JSP:s
and EJB:s. We've load tested on many different setups and are getting some
interesting results:

Simulated number of users: 25
Computer setup: NT4 sp6, 1 GB RAM, Pentium III ~600 MHz
Tomcat setup: 1 ajp12-worker.. (We need out-of-process workers for
scalability)


setup1: webserver - apache 1.3.x, jsp - tomcat 3.2b6 (mod_jk.dll)
setup2: webserver - weblogic, jsp - weblogic (latest versions as of this
date)
setup3: webserver - orion, jsp - orion (latest versions as of this date)


setup2 and setup3 are 8 (eight) times faster in our tests (Load Runner).
Even if we suspected tomcat to be a bit slower, this is truly remarkable..
We've tried optimizing Tomcat by:
1. using ajp13 protocol - failed, we're just getting exceptions from tomcat
2. reloadable=false - of course
3. fiddling with JVM memory settings - minor improvements

Any ideas on what's wrong here? This must be a configuration problem. Have
anyone else found the same results?

Regards,
Manne Anliot.

Future deveploment of Tomcat

[Christian Mallwitz]

Hi,

Knowing that Tomcat is an onging public effort and that most the people
working on it are volunteers I would like know if there is something like a
"release schedule". I've seen discussion about Tomcat 3.3 but  3.2 is still
beta (working great although) and 4.0 is available as milestone builds.

Will 3.2 release before Christmas?
What is the plan about 4.0 (Q1 or Q2 2001)?

Thanks
Christian

Requirement of Apache to integrate with Tomcat

[Joseph Wong]

Hi all,

I want to ask what is the version of Apache to which Tomcat 3.1 can be 
integrated?

thanks.
_
Get Your Private, Free E-mail from MSN Hotmail at http://www.hotmail.com.

Share information about yourself, create your own public profile at 
http://profiles.msn.com.

Antwort: RE: tomcat vs orion vs weblogic testing results (Load Runner)

[thomas . maesing]

Hi,
which one is faster, Tomcat or Websphere?

Thomas





David Oxley <[EMAIL PROTECTED]> am 20.11.2000 15:15:02

Bitte antworten an [EMAIL PROTECTED]

An:"'[EMAIL PROTECTED]'" <[EMAIL PROTECTED]>
Kopie:
Thema: RE: tomcat vs orion vs weblogic testing results (Load Runner)


Are there any benchmarks between Tomcat and WebSphere because at our site
we
are seeing Tomcat is twice the speed of Tomcat and I'm sure this can't be
right!!
Dave
[EMAIL PROTECTED]

-Original Message-
From: Amrhein, Thomas [mailto:[EMAIL PROTECTED]]
Sent: 20 November 2000 11:38
To: [EMAIL PROTECTED]
Subject: AW: tomcat vs orion vs weblogic testing results (Load Runner)


Which JVM did you use?


-Ursprüngliche Nachricht-
Von: Manne Anliot [mailto:[EMAIL PROTECTED]]
Gesendet am: Montag, 20. November 2000 11:15
An: '[EMAIL PROTECTED]'
Betreff: tomcat vs orion vs weblogic testing results (Load Runner)

Hi all..

I'm doing work load testing on a webapplication built on javascript, .JSP:s
and EJB:s. We've load tested on many different setups and are getting some
interesting results:

Simulated number of users: 25
Computer setup: NT4 sp6, 1 GB RAM, Pentium III ~600 MHz
Tomcat setup: 1 ajp12-worker.. (We need out-of-process workers for
scalability)


setup1: webserver - apache 1.3.x, jsp - tomcat 3.2b6 (mod_jk.dll)
setup2: webserver - weblogic, jsp - weblogic (latest versions as of this
date)
setup3: webserver - orion, jsp - orion (latest versions as of this date)


setup2 and setup3 are 8 (eight) times faster in our tests (Load Runner).
Even if we suspected tomcat to be a bit slower, this is truly remarkable..
We've tried optimizing Tomcat by:
1. using ajp13 protocol - failed, we're just getting exceptions from tomcat
2. reloadable=false - of course
3. fiddling with JVM memory settings - minor improvements

Any ideas on what's wrong here? This must be a configuration problem. Have
anyone else found the same results?

Regards,
Manne Anliot.

Servlet context

[David Oxley]

We have written a servlet that is being used in a virtual host setup. I need
to retrieve the virtual host that the servlet has been setup up to work
against. I can retrieve the server name but not the virtual host name. Also
I would like to retrieve the context.
We are using Apache 1.3.14 mod_jk'ed with Tomcat 3.2b7 using ajp13.

Dave
[EMAIL PROTECTED]

RE: tomcat vs orion vs weblogic testing results (Load Runner)

[David Oxley]

Are there any benchmarks between Tomcat and WebSphere because at our site we
are seeing Tomcat is twice the speed of Tomcat and I'm sure this can't be
right!!
Dave
[EMAIL PROTECTED]

-Original Message-
From: Amrhein, Thomas [mailto:[EMAIL PROTECTED]]
Sent: 20 November 2000 11:38
To: [EMAIL PROTECTED]
Subject: AW: tomcat vs orion vs weblogic testing results (Load Runner)


Which JVM did you use?


-Ursprüngliche Nachricht-
Von: Manne Anliot [mailto:[EMAIL PROTECTED]]
Gesendet am: Montag, 20. November 2000 11:15
An: '[EMAIL PROTECTED]'
Betreff: tomcat vs orion vs weblogic testing results (Load Runner)

Hi all..

I'm doing work load testing on a webapplication built on javascript, .JSP:s
and EJB:s. We've load tested on many different setups and are getting some
interesting results:

Simulated number of users: 25
Computer setup: NT4 sp6, 1 GB RAM, Pentium III ~600 MHz
Tomcat setup: 1 ajp12-worker.. (We need out-of-process workers for
scalability)


setup1: webserver - apache 1.3.x, jsp - tomcat 3.2b6 (mod_jk.dll)
setup2: webserver - weblogic, jsp - weblogic (latest versions as of this
date)
setup3: webserver - orion, jsp - orion (latest versions as of this date)


setup2 and setup3 are 8 (eight) times faster in our tests (Load Runner).
Even if we suspected tomcat to be a bit slower, this is truly remarkable..
We've tried optimizing Tomcat by:
1. using ajp13 protocol - failed, we're just getting exceptions from tomcat
2. reloadable=false - of course
3. fiddling with JVM memory settings - minor improvements

Any ideas on what's wrong here? This must be a configuration problem. Have
anyone else found the same results?

Regards,
Manne Anliot.

RE: Session timeout

[Ivan E. Markovic]

Thank you very much.

I v a n ...


>Default is 30 minutes and you can configure it per context :
>
> will make it 1 minute in
>the test context.
>
>Wellington Silva
>UN/FAO
>
>   -Original Message-
>   From:   Ivan E. Markovic [mailto:[EMAIL PROTECTED]]
>   Sent:   20 November 2000 15:58
>   To: [EMAIL PROTECTED]
>   Subject:Session timeout
>
>   Quick question. How long are sessions inside Apache/Tomcat
>valid for?
>   In all the Servlet documentation I have it is stated that
>session
>   duration is server specific. So I checked the Tomcat and
>Apache docs
>   and could find no indicator as to what that duration was.
>Anyone able
>   to enlighten me?
>
>   Thank you.
>
>   I v a n ...
>   Ivan Markovic
>   SculptLight
>   http://www.sculptlight.com
>   (+353) 87 2939256
>   (+353) 1 2982205
>
>   114 Lower Churchtown Rd,
>   Dublin 14,
>   Ireland.

Ivan Markovic
SculptLight
http://www.sculptlight.com
(+353) 87 2939256
(+353) 1 2982205

114 Lower Churchtown Rd,
Dublin 14,
Ireland.

RE: Session timeout

[Lacerda, Wellington (AFIS)]

Default is 30 minutes and you can configure it per context :

 will make it 1 minute in
the test context.

Wellington Silva
UN/FAO

-Original Message-
From:   Ivan E. Markovic [mailto:[EMAIL PROTECTED]]
Sent:   20 November 2000 15:58
To: [EMAIL PROTECTED]
Subject:Session timeout

Quick question. How long are sessions inside Apache/Tomcat
valid for? 
In all the Servlet documentation I have it is stated that
session 
duration is server specific. So I checked the Tomcat and
Apache docs 
and could find no indicator as to what that duration was.
Anyone able 
to enlighten me?

Thank you.

I v a n ...
Ivan Markovic
SculptLight
http://www.sculptlight.com
(+353) 87 2939256
(+353) 1 2982205

114 Lower Churchtown Rd,
Dublin 14,
Ireland.

Anyway to render JSP to static HTML? -- What happens with Cocoon XSP servlet examples?

[cH3rN0b|L]

Hi all!
I was wondering if we can retake the topic I read 
sometime ago in the list
talking about saving JSP results to static 
HTML.
I'm hard behind that, but I have no luck with the 
answers.
I'm using Apache 1.3.12, Tomcat 3.2b5 and Cocoon 
1.8 on RedHat 7.0,
working toghether, 
after some hard work to configure it.
I need ,in some cases, JSP results stored 
in static html files
instead of sending the results to the client 
 internet navigator.
 
Another question could be why in Tomcat 3.2b5 isn't 
included the
xml.jar file, that suspectly starts XSP in Cocoon, 
because I couldn't
ever start that XSP servlet examples that comes with Cocoon.
TIA for any answer that could help me.
 
Iván FontánTechnical Department
 
LATINIA CORPORATION www.latinia.com
 
Spain OfficeBarcelonaTrav. Gràcia, 
342-34408025 Barcelona (Spain)Tel.: +34 934 465 026
 
US OfficesNew York885 3rd AvenueSuite 
2800New York, NY 10022
 
Miami1210 Brickell AvenueSuite 
430Miami, FL 33131

mod_jk and Incomplete Read error

[brent . johnson]

I'm getting the following error when Apache attempts to send a servlet
request to Tomcat (running 3.2b):

sendorder
Incomplete read, deal with it 8188 2048
2000-11-20 11:07:21 - Ctx(  ): Exception in: R(  + /servlet/HPMain + null)
- java.lang.NullPointerException
at HPMain.doPost(HPMain.java:96)
at javax.servlet.http.HttpServlet.service(HttpServlet.java:760)

BAD packet 17487
In: : [B@8199 4/26161/8192
44 4f 66 31 05 01 1b 2d 2d 2d 20 64 65 62 75 67  | DOf1...debug
20 70 6f 69 6e 74 20 31 30 20 2d 2d 2d 0a 2d 2d  | .point.10...
2d 20 64 65 62 75 67 20 70 6f 69 6e 74 20 32 30  | ..debug.point.20
20 2d 2d 2d 0a 2d 2d 2d 20 64 65 62 75 67 20 70  | .debug.p
6f 69 6e 74 20 33 30 20 2d 2d 2d 0a 3c 68 31 3e  | oint.30..h1.
45 72 72 6f 72 3a 20 35 30 30 3c 2f 68 31 3e 0d  | Error..500..h1..
0a 3c 68 32 3e 4c 6f 63 61 74 69 6f 6e 3a 2f 73  | ..h2.Location..s
65 72 76 6c 65 74 2f 48 50 4d 61 69 6e 3c 2f 68  | ervlet.HPMain..h
32 3e 3c 62 3e 49 6e 74 65 72 6e 61 6c 20 53 65  | 2..b.Internal.Se
72 76 6c 65 74 20 45 72 72 6f 72 3a 3c 2f 62 3e  | rvlet.Error...b.
3c 62 72 3e 3c 70 72 65 3e 6a 61 76 61 2e 6c 61  | .br..pre.java.la


If I try and access the servlet through the internal Http server on port
8080 it works.  Also - are the load balancing features working in 3.2b?  I
created a load balancing worker in the workers.properties file then
directed the Ajp13ConnectionHandler port to point at the load balancing
worker's port - but then I couldnt get the servlet to work even from the
internal Http port (connection failed error).

Thanks,

- Brent

Re: Strange behaviour of Tomcat/Apache and session management

[Wolfgang Trexler]

"Stubenrauch,Andreas" wrote:
> 
> This is a known bug in Tomcat 3.1
> 
> URL-encoded session work in Tomcat3.2beta5 and should work again in Tomcat
> 3.2 beta8 upwards

Hmmm, are you sure that it's the same bug? If it's a problem with Tomcat
alone, why does the behaviour change if I change the Listen/Port
directive in httpd.conf and restart Apache?

Just vurious: have you tried "Port 80" instead of "Listen x.x.x.x:y" in
beta8, does it work?

best regards
Wolfgang

> 
> Regards,
> Andreas
> 
> > -Original Message-
> > From: Wolfgang Trexler [mailto:[EMAIL PROTECTED]]
> > Sent: Monday, November 20, 2000 3:46 PM
> > To: [EMAIL PROTECTED]
> > Subject: Strange behaviour of Tomcat/Apache and session management
> >
> >
> > Hi,
> >
> > while debugging a WAP project with URL encoded Tomcat Sessions
> > (JSESSIONID) we encountered the following strange behavior:
> >
> > WAP devices are (normally) not capable of using cookies, the session
> > management of Tomcat (if called standalone on port 8080)
> > recognized this
> > and the method HttpServletResponse.encodeURL(url) returns a valid URL
> > with session ID's encoded within the URL.
> >
> > Once we switched to the production environment with Apache and mod_jk
> > (using apjv12) we encountered, that the method encodeURL(url) returned
> > URL's without session ID's encoded. Thus it seems Tomcats session
> > management thinks the device is capable of using cookies. We played
> > around with the setting in Tomcat and Apache and found the following
> > _strange_ behavior:
> >
> > - in httpd.conf the directive
> >   Port 80 (and no Listen directive) works NOT!
> > - in httpd.conf the directive
> >   Listen 80 (and no Port directive) works
> >
> > We didn't change anything else within the whole environment and were
> > able to reproduce this, that means we could switch between "session
> > handling without cookies - working/non-working" by changing the
> > Listen/Port directive and restarting the Apache server.
> >
> > We've the following environment:
> > - Sun Solaris 2.6
> > - Solaris JDK 1.2.2_06_sparc
> > - Apache 1.3.12
> > - Tomcat 3.1 (final)
> >
> > Any comments/ideas on that?
> >
> >
> > btw: thanks for the good work!
> >
> > best regards
> > Wolfgang

-- 
KPNQwest Austria GmbH., A-1150 Wien, Diefenbachgasse 35
Ing. Wolfgang Trexler [EMAIL PROTECTED]
Web-Technik Business Solutions
Tel.: +43 (1) 899 33 - 141, Fax: +43 (1) 899 33 -10 141

Session timeout

[Ivan E. Markovic]

Quick question. How long are sessions inside Apache/Tomcat valid for? 
In all the Servlet documentation I have it is stated that session 
duration is server specific. So I checked the Tomcat and Apache docs 
and could find no indicator as to what that duration was. Anyone able 
to enlighten me?

Thank you.

I v a n ...
Ivan Markovic
SculptLight
http://www.sculptlight.com
(+353) 87 2939256
(+353) 1 2982205

114 Lower Churchtown Rd,
Dublin 14,
Ireland.

RE: Strange behaviour of Tomcat/Apache and session management

[Stubenrauch,Andreas]

This is a known bug in Tomcat 3.1

URL-encoded session work in Tomcat3.2beta5 and should work again in Tomcat
3.2 beta8 upwards

Regards,
Andreas

> -Original Message-
> From: Wolfgang Trexler [mailto:[EMAIL PROTECTED]]
> Sent: Monday, November 20, 2000 3:46 PM
> To: [EMAIL PROTECTED]
> Subject: Strange behaviour of Tomcat/Apache and session management
> 
> 
> Hi,
> 
> while debugging a WAP project with URL encoded Tomcat Sessions
> (JSESSIONID) we encountered the following strange behavior:
> 
> WAP devices are (normally) not capable of using cookies, the session
> management of Tomcat (if called standalone on port 8080) 
> recognized this
> and the method HttpServletResponse.encodeURL(url) returns a valid URL
> with session ID's encoded within the URL. 
> 
> Once we switched to the production environment with Apache and mod_jk
> (using apjv12) we encountered, that the method encodeURL(url) returned
> URL's without session ID's encoded. Thus it seems Tomcats session
> management thinks the device is capable of using cookies. We played
> around with the setting in Tomcat and Apache and found the following
> _strange_ behavior:
> 
> - in httpd.conf the directive 
>   Port 80 (and no Listen directive) works NOT!
> - in httpd.conf the directive 
>   Listen 80 (and no Port directive) works
> 
> We didn't change anything else within the whole environment and were
> able to reproduce this, that means we could switch between "session
> handling without cookies - working/non-working" by changing the
> Listen/Port directive and restarting the Apache server.
> 
> We've the following environment:
> - Sun Solaris 2.6
> - Solaris JDK 1.2.2_06_sparc
> - Apache 1.3.12
> - Tomcat 3.1 (final)
> 
> Any comments/ideas on that?
> 
> 
> btw: thanks for the good work!
> 
> best regards
> Wolfgang
> -- 
> KPNQwest Austria GmbH., A-1150 Wien, Diefenbachgasse 35
> Ing. Wolfgang Trexler [EMAIL PROTECTED]
> Web-Technik Business Solutions
> Tel.: +43 (1) 899 33 - 141, Fax: +43 (1) 899 33 -10 141
> 

Re: Tomcat 3.1 : Servlet URl without "/servlet/"

[Miles Daffin]

Hi Scal,

> I have a servlet called "myServlet"
> and i have defined  a context "foo"
>
> How i can access to this servlet whitout "servlet" in the url ->
> http://locahost/foo/servlet/myServlet

Try doing this in your apps web.xml (in the etc dir??).


   
   myServlet
   
   
   No description as yet.
   
   
  SomeFoolServlet
   


   myServlet
   /myServlet


This works for me - doing this it seems you can get TomCat to deal with
Servlets ANYWHERE.

Someone please tell me if there is something wrong with this or if I have
missed some vital point
(if you have the time...).

> I will prefer to access it with http://localhost/foo/myServlet.

If you do the above this is exactly what will  happen.

Miles

Serving welcome page v. issuing HTTP redirect

[Phil Radden]

There appears to be a change of behaviour between Tomcat 3.1 and 3.2bX
regarding the handling of the top page for a servlet.

I have a servlet in ROOT context, and wish a particular file, say index.jsp,
to be used for requests for 'GET / HTTP/1.0'.  In 3.1, this is what happens.
In 3.2, an HTTP redirect to 'Location: /index.jsp' is issued instead.

Ideally, I'd want the 3.1 behaviour.  And failing that, the redirection
should really be to a fully-qualified URL.  Any suggestions on what I should
try?

Many thanks

Strange behaviour of Tomcat/Apache and session management

[Wolfgang Trexler]

Hi,

while debugging a WAP project with URL encoded Tomcat Sessions
(JSESSIONID) we encountered the following strange behavior:

WAP devices are (normally) not capable of using cookies, the session
management of Tomcat (if called standalone on port 8080) recognized this
and the method HttpServletResponse.encodeURL(url) returns a valid URL
with session ID's encoded within the URL. 

Once we switched to the production environment with Apache and mod_jk
(using apjv12) we encountered, that the method encodeURL(url) returned
URL's without session ID's encoded. Thus it seems Tomcats session
management thinks the device is capable of using cookies. We played
around with the setting in Tomcat and Apache and found the following
_strange_ behavior:

- in httpd.conf the directive 
  Port 80 (and no Listen directive) works NOT!
- in httpd.conf the directive 
  Listen 80 (and no Port directive) works

We didn't change anything else within the whole environment and were
able to reproduce this, that means we could switch between "session
handling without cookies - working/non-working" by changing the
Listen/Port directive and restarting the Apache server.

We've the following environment:
- Sun Solaris 2.6
- Solaris JDK 1.2.2_06_sparc
- Apache 1.3.12
- Tomcat 3.1 (final)

Any comments/ideas on that?


btw: thanks for the good work!

best regards
Wolfgang
-- 
KPNQwest Austria GmbH., A-1150 Wien, Diefenbachgasse 35
Ing. Wolfgang Trexler [EMAIL PROTECTED]
Web-Technik Business Solutions
Tel.: +43 (1) 899 33 - 141, Fax: +43 (1) 899 33 -10 141

Tomcat 3.1 : Servlet URl without "/servlet/"

[Pascal Bourachot]

I have a servlet called "myServlet"

and i have defined  a context "foo"

How i can access to this servlet whitout "servlet" in the url ->
http://locahost/foo/servlet/myServlet

I will prefer to access it with http://localhost/foo/myServlet.

I saw that with Tomcat 3.2b-8, we can use

 className="org.apache.tomcat.request.InvokerInterceptor"
> debug="0" prefix="/servlet/" />

but i use tomcat 3.1.


Thanks

scal.

RE: tomcat vs orion vs weblogic testing results (Load Runner)

[John Bateman]

Hi

Would you be willling to post the results of your tests here so we can see
what, and where, the bottlenecks are?

Thanks


> -Original Message-
> From: Manne Anliot [mailto:[EMAIL PROTECTED]]
> Sent: Monday, November 20, 2000 5:15 AM
> To: '[EMAIL PROTECTED]'
> Subject: tomcat vs orion vs weblogic testing results (Load Runner)
>
>
> Hi all..
>
> I'm doing work load testing on a webapplication built on
> javascript, .JSP:s
> and EJB:s. We've load tested on many different setups and are
> getting some
> interesting results:
>
> Simulated number of users: 25
> Computer setup: NT4 sp6, 1 GB RAM, Pentium III ~600 MHz
> Tomcat setup: 1 ajp12-worker.. (We need out-of-process workers for
> scalability)
>
> 
> setup1: webserver - apache 1.3.x, jsp - tomcat 3.2b6 (mod_jk.dll)
> setup2: webserver - weblogic, jsp - weblogic (latest versions
> as of this
> date)
> setup3: webserver - orion, jsp - orion (latest versions as of
> this date)
> 
>
> setup2 and setup3 are 8 (eight) times faster in our tests
> (Load Runner).
> Even if we suspected tomcat to be a bit slower, this is truly
> remarkable..
> We've tried optimizing Tomcat by:
> 1. using ajp13 protocol - failed, we're just getting
> exceptions from tomcat
> 2. reloadable=false - of course
> 3. fiddling with JVM memory settings - minor improvements
>
> Any ideas on what's wrong here? This must be a configuration
> problem. Have
> anyone else found the same results?
>
> Regards,
> Manne Anliot.

Re: please help me (Apache + modssl + openssl + tomcat)

[Jan Labanowski]

Try: http://www.ccl.net/cca/software/UNIX/apache/


On Mon, 20 Nov 2000, [euc-kr] ¿¬±¸¼Ò ¿£Áø°³¹ßºÎ ¹Ú½ÃÇö wrote:

> hi 
> I want to use Apache + modssl + openssl + tomcat.
> I found next memo during searching intenet .
> ===
> 
> >>If you want to use modSSL (with Apache and Tomcat), I have that 
> >> > running on 
> >> > my machine and can help.  I have a step-by-step doc.  Let me know.
> 
> ===
> Could you give me a step-by-step doc.  for installing Apache + modssl +
> openssl + tomcat..
> I have tried many time but https:// not running..
> I don't know configure httpd.conf for ssl support...
> 
> If it is not bothering you please help me...
> thanks...
> 

Jan K. Labanowski|phone: 614-292-9279,  FAX: 614-292-7168
Ohio Supercomputer Center|Internet: [EMAIL PROTECTED] 
1224 Kinnear Rd, |http://www.ccl.net/chemistry.html
Columbus, OH 43212-1163  |http://www.osc.edu/

unable to start the server

[ravichandran.B]

Dear Sir,
I have installed the tomcat server for standalone system.  But i am unable
to start.  It gives "Invalid ConnectionHandler" Exception when i tried to
start the server.  My operating system is windows NT workstation.  Please do
suggest me in this.

regards
Ravichandran

Re: System properties

[Miles Daffin]

Hi guys,

> > From the looks of things, it is just the way it is, although I was
hoping
> > the there would be some way to put stuff in the web.xml file, maybe in
the
> > servlet tag like:
> >
> > 
> > myproperty
> > 42
> > 
> >
> > Anyone think this is a good idea, or am I talking rubbish?
>
> It's a bad idea.  Keep your web-apps encapsulated.

Why would such a move, if it were possible, break encapsulation?

As long as this property value is never broadcast
i.e. System.setProperty("myProperty", "42");

If you want application local properties then simply create a class to hold
them
with a static method for obtaining them from anywhere in your app (like
System.getProperties()).

Have a Servlet initialize these by passing it the location of a properties
file containing the
props you need as an initialization parameter. This Servlet can be loaded on
startup,
load in the properties, and there you are, unless I have completely
misunderstood.

Check out the API for java.util.Properties#load(InputStream in)

>
> There is two ways to think about a "global" service:
> - providing a service to many servlets, seems like a good thing
> or
> - bleeding out dependencies upon those that don't need them and allowing
> access to those that shouldn't be using it.  bad things
>

What is 'bleeding out dependencies'? Please explain.

> > PS. static, bad?  It is useful! ...
>
> she said and meant _static initializers_ = bad, not static methods, and
> members.  I concur.

Why are static init blocks || methods bad?
Under what circumstances? In a Servlet context?

Please explain.

Miles

AW: tomcat vs orion vs weblogic testing results (Load Runner)

[Amrhein, Thomas]

Which JVM did you use?


-Ursprüngliche Nachricht-
Von: Manne Anliot [mailto:[EMAIL PROTECTED]]
Gesendet am: Montag, 20. November 2000 11:15
An: '[EMAIL PROTECTED]'
Betreff: tomcat vs orion vs weblogic testing results (Load Runner)

Hi all..

I'm doing work load testing on a webapplication built on javascript, .JSP:s
and EJB:s. We've load tested on many different setups and are getting some
interesting results:

Simulated number of users: 25
Computer setup: NT4 sp6, 1 GB RAM, Pentium III ~600 MHz
Tomcat setup: 1 ajp12-worker.. (We need out-of-process workers for
scalability)


setup1: webserver - apache 1.3.x, jsp - tomcat 3.2b6 (mod_jk.dll)
setup2: webserver - weblogic, jsp - weblogic (latest versions as of this
date)
setup3: webserver - orion, jsp - orion (latest versions as of this date)


setup2 and setup3 are 8 (eight) times faster in our tests (Load Runner).
Even if we suspected tomcat to be a bit slower, this is truly remarkable..
We've tried optimizing Tomcat by:
1. using ajp13 protocol - failed, we're just getting exceptions from tomcat
2. reloadable=false - of course
3. fiddling with JVM memory settings - minor improvements

Any ideas on what's wrong here? This must be a configuration problem. Have
anyone else found the same results?

Regards,
Manne Anliot.

RE: tomcat vs orion vs weblogic testing results (Load Runner)

[GOMEZ Henri]

Could we espect too see this benchmark running TOMCAT on 
Unix boxes (Solaris & Sun JDK / Linux & IBM JDK) 

Regards

"La maîtrise des autres n'est que la force. La maîtrise de vous-même est le
véritable pouvoir" 
-- Lao Tzu 
 

>-Original Message-
>From: Miles Daffin [mailto:[EMAIL PROTECTED]]
>Sent: Monday, November 20, 2000 12:20 PM
>To: [EMAIL PROTECTED]
>Subject: Re: tomcat vs orion vs weblogic testing results (Load Runner)
>
>
>> Hi all..
>>
>> I'm doing work load testing on a webapplication built on javascript,
>.JSP:s
>> and EJB:s. We've load tested on many different setups and 
>are getting some
>> interesting results:
>>
>> Simulated number of users: 25
>> Computer setup: NT4 sp6, 1 GB RAM, Pentium III ~600 MHz
>> Tomcat setup: 1 ajp12-worker.. (We need out-of-process workers for
>> scalability)
>>
>> 
>> setup1: webserver - apache 1.3.x, jsp - tomcat 3.2b6 (mod_jk.dll)
>> setup2: webserver - weblogic, jsp - weblogic (latest 
>versions as of this
>> date)
>> setup3: webserver - orion, jsp - orion (latest versions as 
>of this date)
>> 
>>
>> setup2 and setup3 are 8 (eight) times faster in our tests 
>(Load Runner).
>> Even if we suspected tomcat to be a bit slower, this is 
>truly remarkable..
>> We've tried optimizing Tomcat by:
>> 1. using ajp13 protocol - failed, we're just getting exceptions from
>tomcat
>> 2. reloadable=false - of course
>> 3. fiddling with JVM memory settings - minor improvements
>>
>> Any ideas on what's wrong here? This must be a configuration 
>problem. Have
>> anyone else found the same results?
>
>We had similar results here.
>
>There is nothing wrong - Orion, I am reasonably sure, is 
>simply a better
>product.
>The drawback is that it aint free. But it's cheaper than WebSphere, for
>example.
>Tomcat is the cheapest, and when you pay peanuts
>
>Miles
>
>

Re: tomcat vs orion vs weblogic testing results (Load Runner)

[Miles Daffin]

> Hi all..
>
> I'm doing work load testing on a webapplication built on javascript,
.JSP:s
> and EJB:s. We've load tested on many different setups and are getting some
> interesting results:
>
> Simulated number of users: 25
> Computer setup: NT4 sp6, 1 GB RAM, Pentium III ~600 MHz
> Tomcat setup: 1 ajp12-worker.. (We need out-of-process workers for
> scalability)
>
> 
> setup1: webserver - apache 1.3.x, jsp - tomcat 3.2b6 (mod_jk.dll)
> setup2: webserver - weblogic, jsp - weblogic (latest versions as of this
> date)
> setup3: webserver - orion, jsp - orion (latest versions as of this date)
> 
>
> setup2 and setup3 are 8 (eight) times faster in our tests (Load Runner).
> Even if we suspected tomcat to be a bit slower, this is truly remarkable..
> We've tried optimizing Tomcat by:
> 1. using ajp13 protocol - failed, we're just getting exceptions from
tomcat
> 2. reloadable=false - of course
> 3. fiddling with JVM memory settings - minor improvements
>
> Any ideas on what's wrong here? This must be a configuration problem. Have
> anyone else found the same results?

We had similar results here.

There is nothing wrong - Orion, I am reasonably sure, is simply a better
product.
The drawback is that it aint free. But it's cheaper than WebSphere, for
example.
Tomcat is the cheapest, and when you pay peanuts

Miles

Re: help me..

[Miles Daffin]

Hi Thilan,
 
There is good documentation provided with the 
Tomcat download. 
 
%TOMCAT_HOME%\doc
 
Just read it and follow the steps.
 
Miles
 

  - Original Message - 
  From: 
  Thilan 
  Silva 
  To: [EMAIL PROTECTED] 
  Sent: Monday, November 20, 2000 5:44 
  AM
  Subject: help me..
  
  I'm thilan from Sri Lanka
   
  Now i installing tomcat for IIS bcouse i think 
  somting is wrong... so can U tell me how to install and check it working 
  correctly..
  i try to like this 
  my server is david.espada.co.uk
   
  http://david.espada.co.uk/test.jsp
   
  but not working
   
  pls help me.
   
  thank's 
   
  thilan
   
  my mail address is [EMAIL PROTECTED] 
   

RE: System properties

[Stuart Farnan]

I agree that global properties are not ideal, it would be much better to be
able to localise them for each web-application, I thought I would start by
not asking for too much!  From the replies that I have seen so far I will
give the  element a go.  Basically setting a parameter for a
particular web app which I agree means your web apps are less likely to have
name clashes with parameters they set, as is the case with System
properties, hence more encapsulated.

Stu

-Original Message-
From: Jim Rudnicki [mailto:[EMAIL PROTECTED]]
Sent: 20 November 2000 03:11
To: [EMAIL PROTECTED]
Subject: Re: System properties

> From the looks of things, it is just the way it is, although I was hoping
> the there would be some way to put stuff in the web.xml file, maybe in the
> servlet tag like:
>
> 
> myproperty
> 42
> 
>
> Anyone think this is a good idea, or am I talking rubbish?

It's a bad idea.  Keep your web-apps encapsulated.

There is two ways to think about a "global" service:
- providing a service to many servlets, seems like a good thing
or
- bleeding out dependencies upon those that don't need them and allowing
access to those that shouldn't be using it.  bad things

> PS. static, bad?  It is useful! ...

she said and meant _static initializers_ = bad, not static methods, and
members.  I concur.

Jim

RE: tomcat vs orion vs weblogic testing results (Load Runner)

[Jarek Skreta]

Hi there,

What you've got may be consistent with some other benchmarks. Here is the
link to the Orion web site where they publish some of their results.

http://www.orionserver.com/benchmarks/benchmark.html

They put lot of caveats on their test but the trend is similar to yours.

Regds,

Jarek Skreta


> -Original Message-
> From: Manne Anliot [mailto:[EMAIL PROTECTED]]
> Sent: 20 November 2000 10:15
> To: '[EMAIL PROTECTED]'
> Subject: tomcat vs orion vs weblogic testing results (Load Runner)
>
>
> Hi all..
>
> I'm doing work load testing on a webapplication built on
> javascript, .JSP:s
> and EJB:s. We've load tested on many different setups and are getting some
> interesting results:
>
> Simulated number of users: 25
> Computer setup: NT4 sp6, 1 GB RAM, Pentium III ~600 MHz
> Tomcat setup: 1 ajp12-worker.. (We need out-of-process workers for
> scalability)
>
> 
> setup1: webserver - apache 1.3.x, jsp - tomcat 3.2b6 (mod_jk.dll)
> setup2: webserver - weblogic, jsp - weblogic (latest versions as of this
> date)
> setup3: webserver - orion, jsp - orion (latest versions as of this date)
> 
>
> setup2 and setup3 are 8 (eight) times faster in our tests (Load Runner).
> Even if we suspected tomcat to be a bit slower, this is truly remarkable..
> We've tried optimizing Tomcat by:
> 1. using ajp13 protocol - failed, we're just getting exceptions
> from tomcat
> 2. reloadable=false - of course
> 3. fiddling with JVM memory settings - minor improvements
>
> Any ideas on what's wrong here? This must be a configuration problem. Have
> anyone else found the same results?
>
> Regards,
> Manne Anliot.

Re: Servlet invoker woes - servlet-mapping /servlets/* to invoker fails on 3.2

[Rachel Greenham]

On Saturday 18 November 2000 00:29, you wrote:
> As of Tomcat 3.2-b8 (to be created on Monday), you will be able to set a
> parameter in the server.xml file to override the prefix used by the invoker
> servlet.  There is an entry that looks like this:
>
>  className="org.apache.tomcat.request.InvokerInterceptor"
> debug="0" prefix="/servlet/" />
>
> and you can change the prefix attribute to "/servlets/" for your particular
> case.

I already did this, on 3.2b7, the same solution works already. That's how I 
resolved the issue for now, but I wasn't happy with it, because our webapps 
weren't compatible between 3.1 and 3.2, which was worrying to say the least. 
What's needed is a *standard* solution to this problem that will work across 
all API-compliant servlet containers - ie: defined in the WEB-INF/web.xml for 
the webapp. Or I need to know for sure that there is no such standard 
solution and we should be using aliases for all our servlets. I had another 
look through the API specification and to me it looks like the latter, which 
would be a major pain.

> In Tomcat 4.0, the ability to configure global defaults in the
> "conf/web.xml" file is restored to the way it worked in Tomcat 3.1 -- it is
> much more intuitive for things like this.

Because we're trying to write nice well-behaved webapps, we don't want to 
have to change a global setting, only a local-to-webapp setting. I was able 
to do this on Tomcat 3.1.

-- 
Rachel

Tomcat + JNI -- Help!

[Michael Cohen (ETL)]

 Hi,
 
 I have just installed Apache1.3.14 + modssl + tomcat 3.2b5 on 
 Solaris 2.6 and am trying to call a native method from my servlet 
 (this worked fine under JavaWebServer2.0).
 
 The shared library is definitely being loaded (ie no exception is 
 thrown when I tried to load it separately, the lib is on the 
 LD_LIBRARY_PATH etc) but "UnsatisfiedLinkError"
 is still being thrown indicating the implementation of  the 
 native method is not found.
 
 Has anyone come across anything like this?
 
Many thanks,

Mike.

tomcat vs orion vs weblogic testing results (Load Runner)

[Manne Anliot]

Hi all..

I'm doing work load testing on a webapplication built on javascript, .JSP:s
and EJB:s. We've load tested on many different setups and are getting some
interesting results:

Simulated number of users: 25
Computer setup: NT4 sp6, 1 GB RAM, Pentium III ~600 MHz
Tomcat setup: 1 ajp12-worker.. (We need out-of-process workers for
scalability)


setup1: webserver - apache 1.3.x, jsp - tomcat 3.2b6 (mod_jk.dll)
setup2: webserver - weblogic, jsp - weblogic (latest versions as of this
date)
setup3: webserver - orion, jsp - orion (latest versions as of this date)


setup2 and setup3 are 8 (eight) times faster in our tests (Load Runner).
Even if we suspected tomcat to be a bit slower, this is truly remarkable..
We've tried optimizing Tomcat by:
1. using ajp13 protocol - failed, we're just getting exceptions from tomcat
2. reloadable=false - of course
3. fiddling with JVM memory settings - minor improvements

Any ideas on what's wrong here? This must be a configuration problem. Have
anyone else found the same results?

Regards,
Manne Anliot.

Re: Server-Side Includes (was Re: JSP and embedded servlet call)

[GCS]

Hello!

On H, NOV 20, 2000 at 09:45:18 +0100, Kurt Bernhard Pruenner wrote:

> 
 If it's a jsp, then you should use <%@include file="..."%>, which works at
compile time. The  compiles the included page as a
separate file, thus can be changed after the compilation.
The include syntax you specified is just an optional thing. If the web server
support it, okay, but doesn't have to.
Anyway, I want to include output from a servlet. I have found a page for java
ssi at: http://www.areane.com/javahttp://www.areane.com/java
It's free, and it has an exec java method, like exec cgi! The only problem
that you must specify the class with the whole package like: java.io.whatever
Can I reach the servlets this way? I think no, as I didn't specified any
package xxx line in my code. :(

> to include a servlet that does provide the header for most pages... when we
> tried the same thing with some older Tomcat version (I don't know if it was
> 3.0 or 3.1), the above statement didn't do anything at all and was left in the
> page served by Apache. 
 Well, first I was a dumb, and included a <.servlet> tag in my html, without
jserv installed for apache. So that's why it was displayed without being
touched. And the jsp include was work like a forward and not an include (for
me at least).

Hope I gave you some points,
Laszlo

RE: Server-Side Includes (was Re: JSP and embedded servlet call)

[Lacerda, Wellington (AFIS)]

Why don't you use the plain  ?

 (assuming sswstyle
is a subdirectory in your context).

If it is not, then you can try:

<%
  RequestDispatcher rd =
application.getContext("/sswstyle").getRequestDispatcher("/header.jsp");
  rd.include(request,response);
%>

Wellington Silva

-Original Message-
From:   Kurt Bernhard Pruenner [mailto:[EMAIL PROTECTED]]
Sent:   20 November 2000 09:45
To: [EMAIL PROTECTED]
Subject:Server-Side Includes (was Re: JSP and
embedded servlet call)

"Craig R. McClanahan" wrote:
> Boszormenyi Laszlo wrote:
> > How should I implement something like <.!--#exec
cgi="..."--> in a jsp
> > page?
> 
> Tomcat does not support server-side includes via "#exec
cgi".  You can use 
> such commands in HTML pages if you run Tomcat behind a web
server, but not 
> in a JSP page.

There's something I actually wanted to ask for some time; on
our server, we
have had JServ running for quite some time, and we had
absolutely no problems
using



to include a servlet that does provide the header for most
pages... when we
tried the same thing with some older Tomcat version (I don't
know if it was
3.0 or 3.1), the above statement didn't do anything at all
and was left in the
page served by Apache. 

Now, I've tried this again with Tomcat 3.2b7 and the latest
JServ, and while
it at least tries to include something now, I've found that
Tomcat doesn't get
passed the "/sswstyle/header.jsp", but the location of the
page that tries to
include the JSP page (i.e. "/test/test.shtml"), which of
course fails again...
:(

As I assume that this behaviour happens by JServ passing the
wrong location to
Tomcat when invoked from mod_include - is this some planned
behaviour and is
using mod_include with mod_jserv right out of the question,
or is this just a
bug? :/

-- 
Kurt Bernhard Pruenner --- Haendelstrasse 17 --- 4020 Linz
--- Austria
Music: http://www.mp3.com/Leak --- Work:
http://www.ssw.uni-linz.ac.at
...It might be written "Mindfuck", but it's spelt
"L-A-I-N"...
np: Kendall Jackman - Weightless (ambient.01@hyperreal
comp.)

RE: help me..

[Lacerda, Wellington (AFIS)]

Hi Thilan,
 
I'm Wellington from Brazil.
 
I work with Tomcat and IIS. Did you follow the IIS-HowTo deployed with
Tomcat ? What were the results ?
 
Wellington Silva
UN/FAO

-Original Message-
From: Thilan Silva [mailto:[EMAIL PROTECTED]]
Sent: 20 November 2000 05:44
To: [EMAIL PROTECTED]
Subject: help me..


I'm thilan from Sri Lanka
 
Now i installing tomcat for IIS bcouse i think somting is wrong... so can U
tell me how to install and check it working correctly..
i try to like this 
my server is david.espada.co.uk
 
http://david.espada.co.uk/test.jsp  
 
but not working
 
pls help me.
 
thank's 
 
thilan
 
my mail address is [EMAIL PROTECTED]   
 

Server-Side Includes (was Re: JSP and embedded servlet call)

[Kurt Bernhard Pruenner]

"Craig R. McClanahan" wrote:
> Boszormenyi Laszlo wrote:
> > How should I implement something like <.!--#exec cgi="..."--> in a jsp
> > page?
> 
> Tomcat does not support server-side includes via "#exec cgi".  You can use 
> such commands in HTML pages if you run Tomcat behind a web server, but not 
> in a JSP page.

There's something I actually wanted to ask for some time; on our server, we
have had JServ running for quite some time, and we had absolutely no problems
using



to include a servlet that does provide the header for most pages... when we
tried the same thing with some older Tomcat version (I don't know if it was
3.0 or 3.1), the above statement didn't do anything at all and was left in the
page served by Apache. 

Now, I've tried this again with Tomcat 3.2b7 and the latest JServ, and while
it at least tries to include something now, I've found that Tomcat doesn't get
passed the "/sswstyle/header.jsp", but the location of the page that tries to
include the JSP page (i.e. "/test/test.shtml"), which of course fails again...
:(

As I assume that this behaviour happens by JServ passing the wrong location to
Tomcat when invoked from mod_include - is this some planned behaviour and is
using mod_include with mod_jserv right out of the question, or is this just a
bug? :/

-- 
Kurt Bernhard Pruenner --- Haendelstrasse 17 --- 4020 Linz --- Austria
Music: http://www.mp3.com/Leak --- Work: http://www.ssw.uni-linz.ac.at
...It might be written "Mindfuck", but it's spelt "L-A-I-N"...
np: Kendall Jackman - Weightless (ambient.01@hyperreal comp.)