From security points you should at least upgrade to
the latest 1.3 release of apache. (But that has nothing
to do with tomcat, there have been several security
fixes between 1.3.20 and the current version.
I wouldn't change from 1.3 to 2.0 if you feel ok with
the 1.3 family. (I'm quite
Is that the complete statcktrace ?
Isn't there a root cause ?
-Ursprüngliche Nachricht-
Von: Ricardo de Souza Moura [mailto:ricsouzamoura;hotmail.com]
Gesendet: Donnerstag, 17. Oktober 2002 16:15
An: [EMAIL PROTECTED]
Betreff: is this a bug ?
Does anybody knows about this error ?
Because there are some factors that vary with
every page to .
- The size of output created by the caller of forward
- The time the caller needs to create the output
- The time that tomcat needs between calling forward and
performing the check if the response has already
been committed
Note
AFAIK all Sun JDK's 1.4.1 on all plattforms.
-Ursprüngliche Nachricht-
Von: Jhair Tocancipa Triana [mailto:jhair_tocancipa;icon-scm.com]
Gesendet: Freitag, 18. Oktober 2002 13:59
An: [EMAIL PROTECTED]
Betreff: JAVAC leaking memory
Which version (platform) of the JDK is meant
Your setup is not clear to me.
- Do you want to use tomcat and iis as independend
webservers each listenig on port 80 on different IP's ?
- Or do you want to use tomcat behind IIS and just
route requests for jsp's from the IIS to tomcat.
Who is failing to open the socket (tomcat or IIS) ?
As I'm not using tomcat under windows, I can't offer
much direct help, but two tips to verify who is
responsible:
1.)
- configure two instances of tomcat to listen on
port 80 on different IP's.
See if that is possible. If that is possible the
problem is probably with the iis.
Can't tell you the bug ID, but a quote from
(Bhttp://java.sun.com/j2se/1.4.1/changes.html:
(B
(BThe compiler now releases the pointers to its internal
(Bdata structures after compilation completes, so that
(Busing javac inside a java program will not leak space.
(B
(B
(B
No they aren't:
display-name
description
servlet-mapping
-Ursprüngliche Nachricht-
Von: [EMAIL PROTECTED] [mailto:AnnWenzel;aol.com]
Gesendet: Montag, 21. Oktober 2002 18:49
An: [EMAIL PROTECTED]
Betreff: Re: AW: web.xml error
I am having the same problem. My web.xml file is as
Have a close look at the redirect. It looks as if
it redirects to http on the https port.
(Something like http:/somedomain:443
-Ursprüngliche Nachricht-
Von: Filip Sergeys tc [mailto:fsergeys-tc;verzekeringen.be]
Gesendet: Montag, 21. Oktober 2002 13:09
An: Tomcat Users List
Betreff:
The attributes have to be in the specified order.
So you just have to reorder your entries.
servlet
servlet-mapping
session-config
welcome-file-list
resource-ref
-Ursprüngliche Nachricht-
Von: Steltner, Jörn HTC/DE/ESS [mailto:Joern.Steltner;hochtief.de]
Gesendet: Montag, 21. Oktober
That solution is not recommended:
- Connecting for each user counteracts
the use of connection pools. For most
databases connecting to a db takes much
more time than performing a select on a
given connection.
- To implement this you have to implement your
own realm, as the db realm
Have a look in the generated java file in the
work directory that caused this exception
(processviewfiles_dir_4.java) to find the
corresponding location in your jsp.
If you don't see the error post the code
around that line.
-Original Message-
From: Edmund Smith
Not really related to your question, but to your code:
The use of the synchronized as you do it, is not
recommended.
Don't try to optize the synchnized block by wrapping
it in a if. There is no garantee that this will work
as intended. Have a look at the references in:
The best way to avoid the double checke locking issue
is just not to use it.
Before trying to implement 'double checked locking' right,
you have to understand the purpose.
The goal was to reduce synchronized access to achieve more
performance. (Did you have another rationale ?)
In most cases
Maybe this post from last weekend helps you:
-Original Message-
From: W. Egan [mailto:khayman1122;hotmail.com]
Sent: Sunday, October 27, 2002 6:14 PM
To: Tomcat Users List; Nathan Phelps
Subject: Re: Tomcat 4.1.12 taking IPs or Ports not assigned to it?
Its a Microsoft design
The difference is the internal structure.
Ajp13Connector is the the older, more proven
solution but has some architectural problems.
CoyoteConnector is an new connector architecture.
The CoyoteConnector itself is protocol independend
and uses external classes to speak the protocol.
The
I've once posted a skeleton to do that:
One option for you to work around this, would be
(roughly) something like this:
- Create a hashtable that is global to the webapp.
- Create a hashtable for each session
- Store the session hashtable in the application
Hashtable and use the
Have a look at
http://www.hp.com/products1/unix/java/hpjmeter/downloads/license_hpjmeter_1-5.html
(It's just 2 clicks from the start page)
-Original Message-
From: Steinar Bang [mailto:sb;dod.no]
Sent: Wednesday, October 30, 2002 9:10 AM
To: [EMAIL PROTECTED]
Subject: Re: Profiling
window is opened and the url that
is used for the window contains no session id.
Ralph Einfeldt
Uptime Internet Solution Center GmbH
Hamburg, Germany
Hosting, Content Management, Java Consulting
http://www.uptime-isc.de
-Original Message-
From: Mauro Daniel Ardolino [mailto:mauro
4.0 and 4.1 implment the same specs: JSP 1.2, JSDK 2.3
Both specs are backwards compatible with their
predessor.
Nevertheless some things that are not part of the spec
have changed between 4.0 and 4.1.
To know what might cause your problems you should be more
specific what exacly fails with
The short answer is yes.
The longer answer is that we have scripts that sets up
a site and a database with a well defined content (that's
done in less than 2 minutes) from scratch and run our
testcases against that site and database.
-Original Message-
From: Felipe Schnack
It's a complete different tool.
hpjmeter displays profiling data of a vm.
-Original Message-
From: Charles Baker [mailto:rascharles;yahoo.com]
Sent: Wednesday, October 30, 2002 2:50 PM
To: Tomcat Users List
Subject: Re: Profiling with tomcat
I'm late on this thread, so forgive
There is no strict rule to whom a session id belongs.
It depends on the browser and your site.
If you have cookies disabled it's quite easy, whenever
you open a browser or a window with an url that doesn't
contain a session a new session is created. If the url
contains a session id for a valid
optimal on your
combination of os and vm.
Not much chances in this case :(
Ralph Einfeldt
Uptime Internet Solution Center GmbH
Hamburg, Germany
Hosting, Content Management, Java Consulting
http://www.uptime-isc.de
-Original Message-
From: Carlos J. Ramos [mailto:cjramos;genasys.com]
Sent
But keep in mind that it is just an additional option
not a replacement. Mockobjects can't replace the test
for a real write/read check to the database.
-Original Message-
From: jon wingfield [mailto:jon.wingfield;mkodo.com]
Sent: Wednesday, October 30, 2002 4:50 PM
To: Tomcat Users
to present
the result to the user.
- internal
This is the counterpart for the internal indexer.
Ralph Einfeldt
Uptime Internet Solution Center GmbH
Hamburg, Germany
Hosting, Content Management, Java Consulting
http://www.uptime-isc.de
-Original Message-
From: Michele Emmi
control
about the amount of open connections.
Ralph Einfeldt
Uptime Internet Solution Center GmbH
Hamburg, Germany
Hosting, Content Management, Java Consulting
http://www.uptime-isc.de
-Original Message-
From: Reynir Hübner [mailto:reynir;hugsmidjan.is]
Sent: Wednesday, October 30
That feature is caused by the livetime of the session
cookie. That cookie is just valid for the current browser
session. If you close the browser the cookies is deleted.
If you disable cookies you can reenter a session, if
you use an url that contains a valid session id.
-Original
That lines are created by your jsp.
Every linfeed outside of the jsp tags is
reproduced in the output of the page.
If you have the following:
%!
// some code
%
jsp:useBean ... /
jsp:useBean ... /
%
// some code
%
There is one linefeed between each tag,
if you want to avoid that, you
That depends on the editor that you use.
If you have one that doesn't force a line feed
in the last line, that works if you don't make
a line feed on your own.
-Original Message-
From: Power-Netz (Schwarz) [mailto:schwarz;power-netz.de]
Sent: Thursday, October 31, 2002 11:14 AM
To:
- javawrapper.tar.gz
This is a very basic variant of the same technic as in
the openCMS version.
http://htdig.sourceforge.net/files/contrib/wrappers
Ralph Einfeldt
Uptime Internet Solution Center GmbH
Hamburg, Germany
Hosting, Content Management, Java Consulting
http://www.uptime-isc.de
-Original
First of all: your problem hasn't anything to do
with MVC (That's for Model-View-Controller and is
an architecture for the generation of the content,
if this happens in frames or not is independend of
MVC). So better change that topic for further posts.
The best way is to define a inventory map
There is no cookie, the browser deletes the session cookie
when the browser is closed. So tomcat has no chance to
assign the session to this request.
The only portable solution I see would be something like that:
- define an own cookie that survives the closing of the browser.
(have a lookt at
To me it's not clear what you are doing at all.
On which level do you try to connect from linux to win2k ?
- browser - apache
- apache - tomcat
- can you ping the ip of the win2k server from linux ?
- can you ping the name of the win2k server from linux ?
- can you telnet to the name and port on
I think the error message is quite clear:
You have to provide a unique name to your servlet.
E.G.:
servlet-nameMyDefault/servlet-name
-Original Message-
From: François Vallet [mailto:fvallet;infovista.com]
Sent: Thursday, October 31, 2002 3:58 PM
To: '[EMAIL PROTECTED]'
Subject:
Not shure about that, haven't used it.
We have our own form based login that bypasses
the servlet container completly. (As we still
use servlet container that doesnt supplie one)
If nobody else replies you have to try wether you
can use a filter befor j_security_check.
-Original
We use a home grown diconnected resultset to store the
data between request. (That was developed before there
was something like the RowSet available)
Have a look at the following link about a disconnected
RowSet:
Although i'm on your side in this discussion, I have
doubts about one statements: (See below)
-Original Message-
From: Craig R. McClanahan [mailto:craigmcc;apache.org]
Sent: Thursday, October 31, 2002 6:48 PM
To: Tomcat Users List
Subject: RE: JDBC / ThreadLocal pattern.
You get
AFAIK there is a servlet available that is a wrapper
for JNI calls to php.
Search for phpsrvlt.jar or net.php.servlet, it should
be part of the php distribution. (Third or forth hand
knowledge, I don't use php)
Also have a look at
http://www.php.net/manual/en/printwn/ref.java.php#java.servlet
The problem hasn't anythind to do with javac.
Tmcat uses a class from tools.jar to compile the source.
(That's the class that is used by javac itself)
So it's this jar that is not found in tomcats classpath.
This may be caused by the wrong JAVA_HOME as a previous
post said.
-Original
Have look at:
http://httpd.apache.org/docs-2.0/ssl/
-Original Message-
From: Pae Choi [mailto:paechoi;earthlink.net]
Sent: Friday, November 01, 2002 3:27 PM
To: Tomcat Users List
Subject: Re: Apache 1.3 + mod_ssl + mod_jk or mod_jk2 with EAPI under
Windows
mod_ssl is only
Apache 2.0 contains a modified version of mod_ssl.
mod_ssl is part of the binary distribution. (At least in
2.0.43, I don't have prior versions).
From httpd-2.0.40/modules/ssl/README:
MAJOR CHANGES
For a complete history of changes for Apache 2.0 mod_ssl, see the
CHANGES file in the
It should work for the mod_jk. (Havn't tried it myself,
but have seen many posts about it)
Also have a look at http://www.ubeans.com/tomcat
-Original Message-
From: [EMAIL PROTECTED] [mailto:busch;wave-computer.de]
Sent: Friday, November 01, 2002 1:44 PM
To: [EMAIL PROTECTED]
Sorry, made 2 mistakes:
- Didn't see that windows was the plattform
(Little excuse: it was only in the subject of the
mail that I replied to, but not in the body)
- I didn't know that there are differences in the list
of modules between the os'es.
-Original Message-
From: Pae
Can you verify that your JAVA_HOME is used?
If the startup didn't change to much between tomcat
4 and 5 you can do the following:
Start tomcat with
catalina.bat run
this should echo the JAVA_HOME that is used.
-Original Message-
From: Juergen Heckel [mailto:JHeckel;t-online.de]
I
right?
AFAIK the Apache release for WIn32 is different scenario. I do
SSL-enabled Apache is only for Unix/Linux.
I believe that the orginal message was talking about Apache
on Win32 platform.
Correct me if I am wrong.
Pae
- Original Message -
From: Ralph Einfeldt
Looks like /WEB-INF/controls.tld contains a syntax error.
(At least the sax parser thinks tthat there is one)
-Original Message-
From: Juergen Heckel [mailto:JHeckel;t-online.de]
Sent: Friday, November 01, 2002 3:11 PM
To: Tomcat Users List
Subject: Re: [Win32] Tomcat 5.0 -- JavaSDK
We have following reqirements:
- each site can have a different version of a tool
- many sites share the same vesion of the tool
- a site may change the needed version of a tool
- a site may replace a tool by a different one
(switch from postgres to firebird)
We have a setup like this:
, that's housekeeping. As long as you know
who/what has which
file, the fact that there are two copies of the file is
pretty irrelevant
from a practical viewpoint.
John
-Original Message-
From: Ralph Einfeldt [mailto:ralph.einfeldt;uptime-isc.de]
Sent: Tuesday, November 05
Have a look at the access log.
I guess IE 6.0 passes the anchor tag to the request,
the other browsers don't.
Tomcat doesn't know about the special meaning of
the anchor (#123) as it as client side thing.
-Original Message-
From: Schnitzer, Jeff [mailto:JSchnitzer;maxis.com]
Sent:
What is the problem with that mail ?
The mail you are referencing has been sent to the list
roughly a week ago. (I received it through the list
last wednesday)
-Original Message-
From: Matt Fury [mailto:mattfury88;yahoo.com]
Sent: Wednesday, November 06, 2002 5:11 PM
To: Tomcat
For 4.* is not normal, but I can't say if it is normal
for any 3.x version.
To get an idea what causes the increase of threads
you might send kill -QUIT to the vm.
This should produce stacktraces for all threads
in the logfile for stdout or stderr. So you
can at least get an idea what the
For me the link works...
-Original Message-
From: Holger Ebert [mailto:holger;media-engineering.de]
Sent: Thursday, November 07, 2002 11:09 AM
To: Tomcat Users List
Subject: Re: IRIX 6.5 JK Binary Available
this is a dead link, maybe you can have a look
Whether multithreaded will speed this up, depends
on the fact where the time is spent. If 90% of the
time is spent in the cpu then this will not help
much as long as you don't put additional cpu's in
the server.
Optimisation without profiling is worthless. If
you don't know where the time is
Compare your setup against:
http://www.ubeans.com/tomcat/
Make shure that you have set the jvmRoute
in the engine element.
Make shure that you use response.encodeUrl()
for any link and any form action.
Look at the sessionid in the cookie/url. It
should contain information about the route.
What about
ServletContextEvent.getServletContext().getInitParameter() ?
-Original Message-
From: Reynir Hübner [mailto:reynir;hugsmidjan.is]
Sent: Tuesday, November 12, 2002 2:42 PM
To: Tomcat Users List
Subject: ServletContextListener::Initial paramters
I am implementing a
query for that object to show the
content.
As we have implemented something like that,
quite before something like realms came to
the world, I havn't looked at the details
of realms. So I'm not shure how much of the
Realms you can reuse to implement this.
Ralph Einfeldt
Uptime Internet
Some options: (All not very nice)
- run tomcat behind apache.
- rewrite your servlet to implement http1.1
(Have a look at the tomcat source of the tomcat
servlet that serves the files (Can't remember
the name)
- store the blob in the file system and redirect
the browser to that file
mod_jk has three modes of operation
- in process worker
in this case the vm runs inside the memory space
of apache process. This allows for a faster
communication.
- default worker
in this case tomcat and apache are two
distinct processes and communicate through IPC. (
-
For Redhat 7.2 there is a download for 2.0.40 at:
http://www.johnturner.com/howto/apache-tomcat-howto.html
If you can upgrade apache to 2.0.42 there is a binary at:
http://jakarta.apache.org/builds/jakarta-tomcat-connectors/jk/release/v1.2.0/bin/linux/i386/
-Original Message-
From:
That's a problem of your hoster.
If you don't have access to the webserver
configuration you can't change anything on
your own.
Your provider has misconfigured your tomcat
so it's not (correctly) using any of the
connectors. With any of them it's a
configration error if the jsp source is
You can do the following:
- Describe what happened. (Did receive a feedback mail or not)
- Have look at the headers of your mail.
Is the value of Return-Path the address that you used?
-Original Message-
From: Ronald Aronica [mailto:[EMAIL PROTECTED]]
Sent: Tuesday, November 19, 2002
You have different options:
- Use n web applications in one tomcat instance
In this case you just have one CATALINA_BASE
Each web application has its own context.
- Use x instances of tomcat to run 1 to y web
applications per tomcat.
In this case you have to set the environment
Search for error messages in the log files.
This kind of errors typically show themselves in the log.
One thing that may happen is, that your first request was
to short after the startup. Tomcat takes some time to be
ready for serving requests. Until then you may get 500
errors.
Just a loosely cupled bunch of questions/remarks.
Have you configured apache to forward all request to tomcat
or yust /servlet/* and *.jsp ?
To isolate the problem you could create a static page that
contains the same link.
With this page try it with mod_jk enabled/disabled to
see if you can
I think the correct classification is:
tomcat is a J2EE compliant servlet container and jsp engine
That means tomcat implements everything that the J2EE
specs requires from a servlet container and can be
part of a J2EE Server (together with Jonas, jBoss, OpenEJB,
...).
-Original
The reason that this is recommende is that you otherwise
have to include the pot in every link.
The reason why it doesn't work for you, is that ports
below 1024 are restricted to users that have administrative
right on the system.
To use this ports you can do one of the following:
- Run tomcat
Sorry little typo below
pot=port number
-Original Message-
From: Ralph Einfeldt
Sent: Wednesday, November 20, 2002 12:11 PM
To: Tomcat Users List; [EMAIL PROTECTED]
Subject: RE: HTTP and SSL Connector port problems
The reason that this is recommende is that you otherwise
have
Syntaktically that's correct.
This is the way the session id is encoded in url's if no
cookie is there. For requests that are forwarded to tomcat
that's OK as tomcat knows how to deal with that.
But why is the url encoded at all ?
Do you call response.encodeUrl() on images ?
-Original
Message-
From: Thronicke Klaus-Peter [mailto:[EMAIL PROTECTED]]
Sent: Wednesday, November 20, 2002 3:25 PM
To: Ralph Einfeldt
Subject: WG: How do I unsubscribe?
i have the same problem
i know my email-address has changed. and i'm registered on the list with the
old one.
now i have
Hey, jMeter is pure java. So it runs under any os
for that there is a jdk.
-Original Message-
From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED]]
Sent: Wednesday, November 20, 2002 1:27 AM
To: [EMAIL PROTECTED]
Subject: Where is JMeter for Linux/FreeBSD ?
Hi!
Cannot seem to
Maybe you should provide the answer to following questions:
- Do you run tomcat stand alone ?
- If yes which JSSE do you use ?
- Which jvm do you use ?
- On which operating system ?
Maybe you should try to profile tomcat with a tool like
OptimizeIt to see if it is tomcat or the underlying ssl
If a user doesn't use cookies and you don't use
response.encodeUrl() then each request for a jsp will
create a new session that will be killed after the
session timeout.
If you don't need sessions you can use
%@ page session=false %
in each jsp. Otherwise you should encode all links in
your
It sounds as if not the number of threads is the problem,
but the amount of memory that is used by the threads.
May be you should track the memory usage to verify that
the memory is low when this Exception happens)
(Just log the results of Runtime.[free|total]Memory())
-Original
I don't think that you are right.
To run tomcat as root means you have less security than having
more security.
You have to be aware that you open a additional potential
security hole for the user that runs tomcat. (That is not
specific to tomcat, that is true for any application)
E.G.: If
As I said there are at least three options:
- Use tomcat behind a webserver (Apache, IIS) and connect it
with mod_jk*
This is the best documented attempt.
- Use tomcat behind a proxy (Apache, squid, ...).
That means tomcat is still listening on 8080 and the proxy
directs the requests from
What makes you think that you have mod_jk 1.3 or 1.4.
Is is possible that you took the version number of the protocol
that is used by mod_jk? (AJP has version numbers of 1.3 and 1.4)
-Original Message-
From: Brian McCallister [mailto:[EMAIL PROTECTED]]
Sent: Thursday, December 05,
Why do you want to do that ?
I have already posted today 3 solutions to the problem.
Unless you don't have a good reason why none of this works
for you, I don't think anybody will work on that topic.
From an earlier post of mine:
- Use tomcat behind a webserver (Apache, IIS) and connect it
I prefer to use a didicated user (like tomcat)
to give him the just the rights that are needed
to run tomcat and the application.
If there is more than one application using the
user nobody this user starts to get to much rights
in mosts cases.
Explanation:
To run an application under a
Under linux/unix a little 'performace bost' can be
gained by using the cryptpw command directly instead
of using the pearl function.
BTW: Unless it's the main task of the application,
performances isn't the most important thing. To
justify the additional effort to implement that as
jni, the
Here's a pure Java implementation:
http://locutus.kingwoodcable.com/jfd/crypt.html
and there too:
http://manticore.2y.net/Java/examples/
Havn't tried any of them, we call the cryptpw
command from java.
-Original Message-
From: Galbayar Dorjgotov [mailto:[EMAIL PROTECTED]]
Sent:
If you use jdk 1.4 or jsse (don't know if that is part of
the current tomcat distribution), I guess that java.net.URL
finds sun's HttpsURLConnectionImpl before your for the iaik
class/jar.
Either make shure that your lib is found before those
or just use sun's implementation.
(See:
http://java.sun.com/products/jndi/
-Ursprüngliche Nachricht-
Von: Power-Netz (Schwarz) [mailto:[EMAIL PROTECTED]]
Gesendet: Freitag, 28. Juni 2002 11:41
An: Tomcat Users List
Betreff: AW: JNDI - What resources are loaded?
OT: what is JNDI ???
--
To unsubscribe, e-mail:
If you are using the standard jdbc-odbc bridge,
your database has to be on the maschine that
runs tomcat. My guess: if you test local you have
a local database if you test on the server the
database is in a different maschine than tomcat.
-Ursprüngliche Nachricht-
Von: [EMAIL
Who does the 'JSP Generator failure' look like ?
-Ursprüngliche Nachricht-
Von: tolga tunca [mailto:[EMAIL PROTECTED]]
Gesendet: Montag, 1. Juli 2002 14:30
An: [EMAIL PROTECTED]
Betreff: Using Custom JSP Tags
JSP files work good under Tomcat Root, but i couln't
get them working
Have you tried to set the content type with the page directive ?
%@ page contenttype= %
Your example doesn't contain something like that.
-Ursprüngliche Nachricht-
Von: Simon Juden [mailto:[EMAIL PROTECTED]]
Gesendet: Montag, 1. Juli 2002 17:52
An: '[EMAIL PROTECTED]'
Betreff:
I'm surprised that it worked in prior version without that.
According to the spec every thing outside of the jsp tags
is ignored, so this must have been a bug, not a feature.
-Ursprüngliche Nachricht-
Von: Simon Juden [mailto:[EMAIL PROTECTED]]
Gesendet: Montag, 1. Juli 2002 18:19
Other things I noticed with postgres:
- If you want 24x7 uptime the vacuum command is a show
stopper because it optains table locks. So you can't
make updates while the vacuum runs. (Otherwise you risk
deadlocks)
The latest postgres version has shortened the time
where the
If there is no error message in the logs, I guess
it's a deadlock in the database.
Can you provide more info about the jsp that hangs ?
Try a kill -QUIT tomcat pid and have a look at the
stacktrace to see where the threads are.
Have a look at the oracle console (can't remember the name
of the
AFAIK the main error is in the jdk 1.3 and 1.4.
jasper2 just contains (beside other changes) a workaround for
this bugs.
Have a look at this links: (Require login)
http://developer.java.sun.com/developer/bugParade/bugs/4414162.html
I think you have to (stress-)test which combination is best for your
environment and application.
I would test them in this order:
- jdk1.2 + tc4.0
- jdk1.4 + tc4.0 + jasper2
- jdk1.4 + tc4.1
This is more a feeling than a rational decision driven by facts, as
we don't have tomcat in
config.xml ? Thought it should be server.xml.
http://jakarta.apache.org/tomcat/tomcat-4.0-doc/jndi-resources-howto.htm
l
-Ursprüngliche Nachricht-
Von: Jim Mangione [mailto:[EMAIL PROTECTED]]
Gesendet: Dienstag, 2. Juli 2002 17:40
An: [EMAIL PROTECTED]
Betreff:
Tomcat itself can't create a sig 11.
The only ways a sig 11 can happen are errors in the jvm
or in a native library. Tomcat has no native library,
so typical sources for this kind of trouble are the vm,
the native libraries of the jdk or the native libraries
of third party extension (like
If you don't use jsp's or deliver the jsp's precompiled (jspC)
with your application you don't need tools.jar.
If you need tools.jar you don't have to distribute a jdk,
it's enogh to use a jre + tools.jar.
AFAIK sun has changed the license to enable this kind of setup.
-Ursprngliche
What I was thinking about is something like
Thread 1: lock resource A
(Resource can be a certain row in the database)
Thread 2: lock resource B
Thread 2: lock resource A - Waiting for Thread 1
Thread 1: lock resource B - Waiting for Thread 2
-- Deadlock
If you kill tomcat all lock are
You have to compile your servlets with the -g option.
If you still don't get line numbers you may have to disable
the jit/hotspot compiler.
-Ursprüngliche Nachricht-
Von: Marc Logemann [mailto:[EMAIL PROTECTED]]
Gesendet: Mittwoch, 3. Juli 2002 10:53
An: [EMAIL PROTECTED]
Betreff:
Just set the path in server.xml for your webapp to .
Context path= ... /
-Ursprüngliche Nachricht-
Von: Matthew Oatham [mailto:[EMAIL PROTECTED]]
Gesendet: Mittwoch, 3. Juli 2002 15:14
An: '[EMAIL PROTECTED]'
Betreff: Default Web application
How do I get my web application to
If you are using tomcat standalone I don't know a solution
where you can just setup tomcat t do this.
Never the less there are several ways to achieve this goal:
- Create and define a Filter that redirects all request
to a resource to another resource by using
response.sendRedirect()
-
It's not possible to do that.
This is by design, not a bug.
A dirty hack might do the trick:
Write a servlet filter that sets the value
if it is not set. Configure this filter
to be triggered, whenever j_security_check
is requested. (If that's realy possible I
don't know)
- Do you use mod_jk or mod_jk2 ?
- Which tomcat version do you use ?
- What error do you observe ?
-Ursprüngliche Nachricht-
Von: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED]]
Gesendet: Mittwoch, 3. Juli 2002 17:44
An: [EMAIL PROTECTED]
Betreff: VirtualHost
This would mean I have
1 - 100 of 1242 matches
Mail list logo
