Hi everybody.
Someone has the architecture of Tomcat and a brief explanation of this?
For architecture t think at some like this:
http://www.javaportal.it/images/tesi/bonzagni2/Bonzagnic_html_m14d04b4d.png
thanks
--
Stefano
Hi all,
I know this topic, or at least bits and pieces of it, has been covered in
various posts, but just for clarification in my mind I would like the
opinions of experienced Tomcat and Unix/Linux experts about optimizing
performance and security configuration for Tomcat.
As a starting point
Hi!
Why is session unbound called after cluster failover?
Two more questions:
-- In a scenario with two nodes, if I shutdown one, the other is instantly
ready to receive and respond to requests, or it is necessary to wait a
while?
In a procuction scenario we cannot wait, because requests are
En l'instant précis du 25/02/08 13:51, Dave s'exprimait en ces termes:
Our Linux(FC) machine has 8G physical memory and 12G swap size. I am using JDK
1.5. I tried to set the Java option -Xmx to set max heap size for best
performance, the allowed max heap size is 2048M . Does that mean
Hi everybody,
I am planning to set up a couple (2-5) of webapps. I've encountered some
problems in the past when running multiple webapps on a single tomcat,
cause they somehow interfered with each other (frankly, I never
unterstood why).
My question is:
Is there a best practice for running
Hi everyOne
From adobe Faq
On 32-bit processor machines, the largest contiguous memory address
space the operating system can allocate to a process is 1.8GB. Because
of this, the maximum heap size can only be set up to 1.8GB. On 64-bit
processor machines, the 1.8 GB limit does not apply, as
Hi all,
I want to deploy war file in vertical cluster. I'm confused a little
bit with configuration.
I have three clustered tomcats -labeled as (A,B,C)- in same machine.
I have put the
Deployer
className=org.apache.catalina.cluster.deploy.FarmWarDeployer
Dave wrote:
Our Linux(FC) machine has 8G physical memory and 12G swap size.
Happy swapping ;-)
--
Mikolaj Rydzewski [EMAIL PROTECTED]
smime.p7s
Description: S/MIME Cryptographic Signature
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
Matthias,
MK wrote:
| Christopher Schultz chris at christopherschultz.net writes:
|
| How did you set the [Java] property?
|
| I am running my web app inside an Eclipse WTP server container
project. So I set
| the property as a JVM argument in the
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
Andrew,
Andrew Hole wrote:
| Why is session unbound called after cluster failover?
What value is being unbound?
| Two more questions: -- In a scenario with two nodes, if I shutdown
| one, the other is instantly ready to receive and respond to
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
Matthias,
Christopher Schultz wrote:
| From the bug you mentioned in your original post, it looks like
| putting it into your webapp just isn't going to work. Instead, you'll
| have to install it into a ClassLoader that is higher-up in the chain.
|
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
Matthias,
MK wrote:
| Actually I don't even call URL.openConnection, because I don't need it
at all.
| It's really just that the java.net.URL constructor requires that there
exists an
| object which implements this behavior just in case someone
To answer my own question, it looks like jfreechart.
http://sourceforge.net/projects/jfreechart
On Mon, Feb 25, 2008 at 8:50 AM, Dan Armbrust
[EMAIL PROTECTED] wrote:
Does anyone happen to know what package Lamda Probe uses to create its
graphs? They look really nice.
I would search
On Mon, Feb 25, 2008 at 5:46 AM, Abid Hussain [EMAIL PROTECTED] wrote:
My question is:
Is there a best practice for running several webapps on a single
tomcat instance ...
Tomcat is made for this. Install your webapps per the documentation
(and Servlet Spec) and they won't interfere with
A couple of issues:
We've set our session expiration to 12 hours (I know it's long) and
we're seeing behavior where certain browsers (namely IE) apparently
can't count that high (we set the meta Refresh header but the page
doesn't reload after the allotted time, session expiration time + 20
Hi,
in order to start Bests practises, i think about these rules...
Libraries
- Be care of [Tomcat directory]\shared\classes or [Tomcat
directory]\shared\lib contents
- Be care of [Tomcat directory]\common\classes, [Tomcat
directory]\common\classes\endorsed and [Tomcat
Hi!
Well, our class USER implements HttpSessionBindingListener,
Serializable and when failover happens valueUnbound is executed(public
void valueUnbound(HttpSessionBindingEvent event)) - valueUnbound is a
method of class USER.
Could you help.
Thanks
On 2/25/08, Christopher Schultz [EMAIL
Hi,
I have a tomcat server running a lone java application on vms. The problem is
that it shuts down occasionally for seemingly no reason. Could somebody help
shed some light on the following log snippet - it's the entry right before the
shutdown happens. I'm really a novice when it comes to
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
Andrew,
Andrew Hole wrote:
| Well, our class USER implements HttpSessionBindingListener,
| Serializable and when failover happens valueUnbound is executed(public
| void valueUnbound(HttpSessionBindingEvent event)) - valueUnbound is a
| method of
I installed Linux FC6 64-bit on the machine DELL 2590(I think it is INTEL type
CPU). But JVM 64-bit is only available for AMD and SPARC. Is the SUN not
support INTEL?
Thanks, Dave
David Delbecq [EMAIL PROTECTED] wrote:
En l'instant précis du 25/02/08 13:51, Dave s'exprimait en ces
I have an ear application with entity,business and web modules.
application ...
module
ejbfoo.par/ejb
/module
module
ejbbar.ejb3/ejb
/module
module
web
web-uribaz.war/web-uri
context-root//context-root
/web
/module
Hi,
We've set our session expiration to 12 hours (I know it's long) and
we're seeing behavior where certain browsers (namely IE) apparently
can't count that high (we set the meta Refresh header but the page
doesn't reload after the allotted time, session expiration time + 20
minutes).
you do need to define the deployer on each instance,
however in 5.5 there was an unfortunate change, and I believe that
deployDir has to point to your webapps directory
Filip
Taner Diler wrote:
Hi all,
I want to deploy war file in vertical cluster. I'm confused a little
bit with
Hello,
cookie.setVersion(1) remembers the cookie only for the browser session. A
new browser does not have access to the cookie
We did cookie.setMaxAge(Integer.MAX_VALUE) but that doesn't help.
Any thoughts please?
Thanks,
Sushil Vegad
Technical Lead, Scheduling Project
Serebrum Corporation -
Thanks for help. So am I right saying that no conflicts between webapps
can occur when
1. using tomcat 5.x...
- all wepapps only use their WEB-INF/lib as directory for libraries and
- none of the webapps uses libraries from $CATALINA_HOME/common/lib
resp. $CATALINA_HOME/shared/lib
2. using
For some reason Linux always calls 64 bit OSes 'AMD' - in fact, the sun
64 bit AMD version works fine on modern 64 bit Intel CPUs. The confusion
comes because there was an older 64 bit design from Intel called the
'Itanium' which was intended for servers and had a completely different
It happens on both. Valueunbound is executed on old server when tomcat
is stopped, and is executed on new server when I execute logout (or if
I shutdown tomcat). We only want that new server executes
valueunbound.
Could you help?
Thanks a lot
A
On 2/25/08, Christopher Schultz [EMAIL PROTECTED]
Paul-
Are you saying that certain browser will never expire their sessions?
Or are you saying that certain browsers kill their sessions before 12
hours (because they can't count that high)?
The former, i.e. that browsers will never expire the sessions.
The way I understand it - you are
On Fri, Feb 22, 2008 at 12:57:22PM -0800, Alan Chaney wrote:
Or as I mentioned in a recent email, you can run something like jsvc and
set the user to 'tomcat' which allows you to bind to the port and then
changes the user.
Okay, either I wasn't paying attention the last time I looked at jsvc,
the stack trace idea is a good idea, and you can post it here
Filip
Christopher Schultz wrote:
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
Andrew,
Andrew Hole wrote:
| Well, our class USER implements HttpSessionBindingListener,
| Serializable and when failover happens valueUnbound is
I'm trying to create a Realm that authenticates a user and logs the IP address
they attempted from, but I am having trouble finding how I get the IP address
of the request (the Remote IP Address).
I know I could get it if I had the Socket object, but I don't see how I get
that from the Realm.
My application has in META-INF/context.xml:
Context cachingAllowed=false /
This works just beautifully in tomcat-6 but in tomcat-5.5 the
application warfile fails to unpack.
Not a single related log line is written in any of tomcat's logfiles!
Any idea what's up with tomcat-5.5?
Any
Note that the user web application mechanism only recognizes one
webapp per user.
If you need more than that, you might write up a little gadget (to be
run setuid) which collects and vets enough information to construct a
context descriptor file and then deposits it in
Classes shared between webapps go in shared/ folder. server/ folder is
for server specific (invisible to webapps) classes.
PS: that information is for tomcat, since you seem to use jboss, you
better ask jboss mailing list, as jboss might use different folders
Dave a écrit :
I have an ear
Well, IF you are sure you want to use Apache HTTPD to frontend Tomcat,
there's no reason for Tomcat to run an HTTP connector at all. Just
use the AJP connector, and only allow it to listen to your frontend
service(s), or firewall it away from the Big Bad Internet. (Or both.)
--
Mark H. Wood,
Stack trace of old server:
java.lang.Exception
at mni.core.sessions.User.valueUnbound(User.java:95)
at
org.apache.catalina.cluster.session.DeltaSession.removeAttributeInternal(DeltaSession.java:1589)
at
On Thu, Oct 25, 2007 at 12:19 PM, Filip Hanik - Dev Lists devm wrote:
browser don't work the way you might it expect to, firefox for example,
will not display anything until the entire request is complete. so the
chat example is no good that way.
write a client application for your
Yes, David, you are correct.
After some reading, I realized that configuration was recommended by my
application's vendor and is indeed erroneous. Thank you for the
clarification explanation.
~LZM~
-Original Message-
From: David Smith [mailto:[EMAIL PROTECTED]
Sent: Sunday, February
Martin-
We are using Struts, however, version 1.2.9. But, after looking at the
link, I'm not sure this will help as it doesn't really address the
problem. Storing the date/time a user logs in on the session is
probably useful, but our problem is that we want to forcefully log the
user out
--- 10:10AM Mon 25 Feb 2008, Adam Gordon
[EMAIL PROTECTED] wrote:
Storing the date/time a user logs in on
the session is
probably useful, but our problem is that we want to
forcefully log the
user out if there's no human present at the computer
and the AJAX tasks
keep a user's session
Kerrin Hardy wrote:
I'm trying to create a Realm that authenticates a user and logs the IP address
they attempted from, but I am having trouble finding how I get the IP address
of the request (the Remote IP Address).
I know I could get it if I had the Socket object, but I don't see how I get
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
Andrew,
Andrew Hole wrote:
| It happens on both. Valueunbound is executed on old server when
| tomcat is stopped, and is executed on new server when I execute
| logout (or if I shutdown tomcat).
That behavior seems entirely consistent with what
-Original Message-
From: Adam Gordon [mailto:[EMAIL PROTECTED]
Sent: Monday, February 25, 2008 13:11
To: Tomcat Users List
Subject: Re: Session expiration and AJAX issues
Martin-
We are using Struts, however, version 1.2.9. But, after
looking at the
link, I'm not sure
Greetings list.
For some reason web browser queries are not passing through apache to tomcat
via mod_jk. I understand I may have a misconfiguration somewhere, I just
cannot find it. L
I am trying to host 4 separate domains, all with different java apps. Each
domain has its own IP and own
I use Apache Tomcat (5.5.9) to host the website (xyz.com) along with a
couple of sub-domains (photos.xyz.com and documents.xyz.com). You can find
the relevant portion of server.xml file below.
Hosting the websites in this way has lead to two main headaches.
1. I cannot share my Java code between
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
Adam,
What you need is to make a request without touching the session.
Tomcat cannot do this by itself; you're going to have to either hack
Tomcat to add a no-touch-session parameter to the session manager
(which wouldn't be a bad TC enhancement
Hi,
I am a student and doing a research on open source processes. I would be
thankful if some one can answer few simple questions to help me in my research.
1. If someone wants to work on a feature or bug, is he/she required to get
permission before working? Is he/she REQUIRED to join the
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
Andrew,
Andrew Hole wrote:
| I think you need to manage your events in a way that takes into account
| the fact that valueUnbound will can be called by other members in the
| cluster.
| It is not a easy task, rigth?
| Could you be more specific in
---
HARBOR: http://coolharbor.100free.com/index.htm
The most powerful application server on earth.
The only real POJO Application Server.
Making the Java dream come true.
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
Kerrin,
Kerrin Hardy wrote:
| I'm trying to create a Realm that authenticates a user and logs the
| IP address they attempted from, but I am having trouble finding how I
| get the IP address of the request (the Remote IP Address).
This is not
I think you need to manage your events in a way that takes into account
the fact that valueUnbound will can be called by other members in the
cluster.
It is not a easy task, rigth?
Could you be more specific in the way of developing it?
Thanks
On Mon, Feb 25, 2008 at 7:27 PM, Christopher Schultz
Well, valueunbound is executed on both TC.
we are doing session replication, lets say we have a 5 (for example) node
cluster. When a user logs on, his/her session will be replicated on 5
tomcats. When he/she logs out, you only want track that the user has logged
out once, not 5 times (our logout
Not sure if I followed this thread closely enough. It looks like
sometimes you are talking about node shutdown (your example stack), and
sometimes about user logout (your last message).
Concerning user logout (resp. session expiration/invalidation): the
attribute notifyListenersOnReplication
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
Andrew,
Andrew Hole wrote:
| Well, valueUnbound is executed on both TC.
You mentioned that. The valueUnbound is being propagated from the TC
instance shutting down across the cluster, right? I can't seem to get a
straight answer to that question.
Are there any plans to have Tomcat handle session replication on a
per-webapp basis?
We currently have several servers configured for session replication
because of a single application. Not only do we not want session values
from other applications on the machines to be replicated as it is
On Feb 25, 2008, at 15:35 , Uzma Khawaja wrote:
Hi,
I am a student and doing a research on open source processes. I
would be thankful if some one can answer few simple questions to
help me in my research.
1. If someone wants to work on a feature or bug, is he/she required
to get
Christopher Schultz schrieb:
Of course, bringing down one container in a cluster will have the effect
of logging everyone out at once, even if they are on other servers as well.
No, unless you set expireSessionsOnShutdown which is false by default,
shutdown of a node should not expire the
Not sure what you mean by this: session replication is local to webapps.
After configuring the cluster element in server.xml, you need to
activate replication for each webapp with a distributable element in
web.xml and only session changes for those webapps with distributable
set get
Adding a time decay in our timer task is an interesting idea and were it
not for IE's JavaScript counting ineptness, that'd probably work.
Bob Hall wrote:
--- 10:10AM Mon 25 Feb 2008, Adam Gordon
[EMAIL PROTECTED] wrote:
Storing the date/time a user logs in on
the session is
probably
Uzma Khawaja wrote:
Hi,
1. If someone wants to work on a feature or bug, is he/she required to get
permission before working? Is he/she REQUIRED to join the community?
No permission required. By very virtue of downloading the software and
working on the bug they have joined the community. It
Since your 3 webapps need to share classes code, object instances and
session information, i don't see why you want to make 3 webapps instead
of just one. So my advise would be to make a single webapp.
Kristin Coles a écrit :
I use Apache Tomcat (5.5.9) to host the website (xyz.com) along with
this has been done since the beginning of session replication.
if your webapp has distributable/ in web.xml sessions will be replicated.
if the element is missing, session data will not be replicated even if
the server is configured for clustering
Filip
Andrew R Feller wrote:
Are there any
Thanks for the reply, David. Are you saying that I need to get rid of the
sub-domains? i.e. I should use xyz.com/photos rather than photos.xyz.com?
Can you please elaborate?
Thanks,
Kristin
On Mon, Feb 25, 2008 at 4:04 PM, david delbecq [EMAIL PROTECTED] wrote:
Since your 3 webapps need to
The valueUnbound is NOT being propagated from the TC instance shutting
down across the cluster. The valueunbound is executed when I do the shutdown
of an TC instance (session logout tracking was done - database record
inserted). After that I still work with this session on the active node ...
what is your scenario,
the following code worked for me, even though the browser doesn't send
up cookie version
%
javax.servlet.http.Cookie[] cs = request.getCookies();
String value = null;
for (Cookie co : cs) {
if (test.equals(co.getName())) value = co.getValue();
}
Hello everyone,
We have a high load environment where we are running tomcat 5.5.15
successfully. We are interesting in reducing the system CPU load and
switching to Tomcat 6 with NIO, but so far have run into a few issues.
After trying out the patch Filip recommended
Actually I've seen something like this. I'm under a lot of pressure to
get something out so I haven't investigated it further, but what seemed
to be happening is that if the client slowed down the NIO connector was
throwing a SocketTimeout. I tried messing about with the timeout
settings but
you don't need four (4) acceptor threads, turn that to one (1)
also, you omited the most interesting stack trace :)
Filip
Emile Litvak wrote:
Hello everyone,
We have a high load environment where we are running tomcat 5.5.15
successfully. We are interesting in reducing the system CPU load
Alan Chaney wrote:
Actually I've seen something like this. I'm under a lot of pressure to
get something out so I haven't investigated it further, but what seemed
to be happening is that if the client slowed down the NIO connector was
throwing a SocketTimeout. I tried messing about with the
Christopher, thanks a lot for your replies.
Christopher Schultz chris at christopherschultz.net writes:
MK wrote:
| Actually I don't even call URL.openConnection, because I don't need it
at all.
| It's really just that the java.net.URL constructor requires that there
exists an
| object
MK kaeppler.matthias at nts.ricoh.co.jp writes:
[...]
I am using Tomcat 5.5 by the way.
I will try rolling a jar from my protocol handler and put it in the places you
suggested. Maybe that'll work, I'll report back.
actually I just realized that this is not gonna help if Tomcat does not
Jason-
the url patterns depends on the servlet mapping you've defined..
You're going to need some algorithm which defines infinity
Your current mechanism binds the current Servlet to the ajax control (div?)
Acquire the initial date/time ..store it somewhere..hopeully in session
when the servlet
--- 2:09PM Mon 25 Feb 2008 Adam Gordon
[EMAIL PROTECTED] wrote:
Adding a time decay in our timer task is an
interesting idea and were it
not for IE's JavaScript counting ineptness, that'd
probably work.
The server could track the requests and provide an
updated delay time for the timer
Make sure cookies are enabled. If they are:
There are two options (1) that it is indeed the same session and that
there is a huge Tomcat bug (2) that these are two separate sessions
but that you are somehow mixing the data - local bug.
To determine which is it, you could do a toString() on
74 matches
Mail list logo
