David Rees schrieb:
So in case the remote host is dead (i.e. it's not only Tomcat not
answering or no Tomcat there), we have the problem, that TCP as a
reliable problem tries hard to establish a connection with several
resends of SYNs in increasing intervals, leading to long waiting times.
On 8/23/07, Christopher Schultz [EMAIL PROTECTED] wrote:
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
Duncan,
Not to belabor this thread too much, but...
Lyallex wrote:
I never actually suggested [curl] was a
hacking tool
See the subject line.
Yes ... fair point :-}
Thanks for all the replies about this question.
But I still can not know if it could be solved by tuning the attributes
in worker.properties.
I read that doc carefully and make some changes to the worker.proerties.
There is no improvement.
It seems that mod_jk try to establish a TCP connection
Guofeng Zhang wrote:
Thanks for all the replies about this question.
But I still can not know if it could be solved by tuning the attributes
in worker.properties.
I read that doc carefully and make some changes to the worker.proerties.
There is no improvement.
It seems that mod_jk try to
Hi Chris,
I guess you're question pointed to Sam, right?
Anyways, that's also what I think.
According to the specs a *noncaught* exception should trigger the errorpage
At least here it's working as expected ;)
Cheers
Gregor
--
what's puzzlin' you, is the nature of my game
gpgp-fp:
Because I have several applications which share the same dataource.
David Smith-2 wrote:
Ok... this begs the question: Why did you add the datasource to
$TOMCAT_BASE/conf/context.xml? The datasource should be in your
individual webapp's context.xml file where all you have to do is
Then I don't think you can avoid a restart of tomcat. The only other
method of defining a jdbc pool shared across webapps involves defining
it in the GlobalNamingResources.../GlobalNamingResources of
conf/server.xml and that also requires a restart.
Pick your poison -- one db pool per webapp
Hello!
What's benefits of using Clustering/Session Replication on same machine?
Thanks a lot
DOH!
Yes, it was a classloading problem. I had assumed the IDE had packaged
commons net in the web app but it hadn't.
MARollins wrote:
Hi,
I've written an application to download and process a set of files using
commons.net FTPClient. This runs fine from in a standalone application.
Hi
Hopefully a simple answer, the documentation says
recovery_options is a bitmask but what I am unsure of is do I set the
value in binary or do I use integer values?
1: don't recover if Tomcat failed after getting the request
2: don't recover if Tomcat failed after sending the headers to
Hello all,
I'm developing a webapp that does lots of ajax requests, but I need to
keep track of the session. Wath i do is to read the JSESSIONID cookie
with javascript and send the value as a POST parameter when I make the
request, but then tomcat doesn't open the session automaticaly. My
question
i want to have a folder that will contain files that can be downloaded/saved
only by going through a servlet.. how can i do this? are there any tutorials
or links? thanks..
--
View this message in context:
http://www.nabble.com/how-to-add-restriction-to-a-folder--tf4323046.html#a12310612
Sent
[EMAIL PROTECTED] schrieb:
Hi
Hopefully a simple answer, the documentation says
recovery_options is a bitmask but what I am unsure of is do I set the
value in binary or do I use integer values?
1: don't recover if Tomcat failed after getting the request
2: don't recover if Tomcat failed
I am not running in a domain, my servers are running in a workgroup
only, and the machine tomcat 5.5 is running on is logged into an
administrator account. It appears that it sees the mapped drives
because if I try to access a file that is not available, I get the
proper JAVA message. When I
I just tried changing the login configuration to log onto the
administrator account, and I still get access denied. Any thoughts???
Thanks, Dan Gross, ATLC
Mark Thomas wrote:
Daniel L. Gross wrote:
I have a servlet that does a direct read from a mapped drive in
Windows. It works fine
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
Aaron,
Don't hijack threads. Please post a new question to the list.
- -chris
Aaron wrote:
I am running Tomcat 5.0.28 and Apache 2.2 using mod_jk. I am constantly
seeing the following error show up in the catalina.log:
WARNING: Exception
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
Brian,
Don't hijack threads. Please post a new question to the list.
- -chris
Brian Barnett wrote:
I have a struts-based application running on multiple tomcat instances, load
balanced by a hardware load balancer, i.e., no Apache Web Server. I
Der Default-Port fuer Tomcat ist definitiv 8080:
http://tomcat.apache.org/tomcat-5.0-doc/deployer-howto.html
Wenn ich Tante Google mal glauben darf, ist der Default Port fuer Squid 3128.
Lediglich in einigen Tutorials wird 8080 empfohlen, wenn man Squid als
HTTP-Proxy einsetzt - ist aber kein
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
Cun,
shunhecun wrote:
Because I have several applications which share the same dataource.
I would configure them each separately in their own META-INF/context.xml
file, even though all the values are the same. Is this feasible in your
environment?
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
Daniel,
Daniel L. Gross wrote:
How do I change the configuration in the tomcat service to allow it to
see other computers?
You don't change the service configuration (unless you run Tomcat as
some other user -- which isn't a bad idea). Instead,
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
Aaron,
Aaron wrote:
Sun v245 running Solaris 10
I think this might have been covered before, but I'm not entirely sure.
You'd have to check the circumstances. Search the archives for
SocketException, socketSetOption, and IPv6. I think that you
Hi
Another easy question to answer, hopefully...
I just spend a lot of time going through this page:
http://tomcat.apache.org/connectors-doc/reference/workers.html
to decide which parameters we need/want to implement.
However I am a bit confused between the
/etc/httpd/conf/workers.properties
Hello,
my Tomcat 6.0.13/14 server is responding in a very long time. It does not
matter if I try to start the main page of the server or one of my webapps.
About 2 minutes for each website. This is not the case, if I test on my
local machine with http://localhost:8080/myapp; (IE or Firefox on a
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
Hezjing,
hezjing wrote:
In Tomcat 5.5\conf\web.xml, I have defined modificationTestInterval=0
and that the context's reloadable=true.
Maybe try 'reloading=true' instead?
http://tomcat.apache.org/tomcat-5.5-doc/jasper-howto.html
- -chris
Thanks for all your help, I found out the problem. It was the service
privilege running under the local machine. The default tomcat icon at
the bottom of the screen that allows you to configure and start/stop
tomcat would not let me change the user it ran under. However, I was
able to
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
Daniel,
Daniel L. Gross wrote:
I was
able to change the user to Administrator
This is basically the same as running as root on a UNIX system: it
works, but nobody can really recommend it since you're giving your app
server full access to
Hi,
after some days of tries and research I was able to run properly tomcat
(5.5.23), apache 2 and mod_jk for the same machine.
So now, my webapps are reachable with both URLs:
http://www.fake.com/webapps
or
http://www.fake.com:8080/webapps
As I understood mod_jk, apache/mod_jk looks in
Hi,
Try removing Connector port=8080 ... element from your tomcat server.xml
Quoting Sebastien Moretti [EMAIL PROTECTED]:
Hi,
after some days of tries and research I was able to run properly tomcat
(5.5.23), apache 2 and mod_jk for the same machine.
So now, my webapps are reachable
try adding
|-Djava.net.preferIPv4Stack=true
to your startup script
Filip
|
Christopher Schultz wrote:
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
Aaron,
Aaron wrote:
Sun v245 running Solaris 10
I think this might have been covered before, but I'm not entirely sure.
You'd have to
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
Andrew,
Andrew Hole wrote:
What's benefits of using Clustering/Session Replication on same machine?
If you have a really buggy application that crashes a lot, clustering
and session replication can buy you some robustness that your engineers
have
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
Gerhardus,
[EMAIL PROTECTED] wrote:
However I am a bit confused between the
/etc/httpd/conf/workers.properties for apache and
/usr/share/tomcat/conf/workers.properties
You only need one workers.properties file: it configures mod_jk for
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
Rapthor,
Rapthor wrote:
But after 1 day the same issue reappeared (response times of about 2
minutes).
When you observe the lng response time, try doing a thread-dump of
the JVM to see what's going on. You can see which threads are waiting on
Right, I understand the function of the reloadable attribute. What I don't
understand is how on a modern computer, registering 2000 locations in
memory/on disk to check for changes could take over a minute. Imagine just
what it is doing, it isn't even checking for changes yet, its just
registering
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
Hernâni,
Don't hijack threads. Please post a new question.
- -chris
Hernâni Henrique Ramos Cerqueira wrote:
Hello all,
I'm developing a webapp that does lots of ajax requests, but I need to
keep track of the session. Wath i do is to read the
Sex, 2007-08-24 às 10:06 -0400, Christopher Schultz escreveu:
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
Hernâni,
Don't hijack threads. Please post a new question.
???
- -chris
Hernâni Henrique Ramos Cerqueira wrote:
Hello all,
I'm developing a webapp that does lots of ajax
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
Hernâni,
I'm developing a webapp that does lots of ajax requests, but I need to
keep track of the session. What I do is to read the JSESSIONID cookie
with javascript and send the value as a POST parameter when I make the
request, but then tomcat
Dan Beaulieu wrote:
Right, I understand the function of the reloadable attribute. What I don't
understand is how on a modern computer, registering 2000 locations in
memory/on disk to check for changes could take over a minute. Imagine just
what it is doing, it isn't even checking for changes
I think that really depends on the path you use inside
your httpd.conf. That's the place where you specify
which workers.properties to use.
Cheers.
--- [EMAIL PROTECTED] escribió:
Cool, it is starting to make a bit more sense.
Does that mean that setting settings in
Well I guess this gets into the nuances of classloaders that I am completely
unfamiliar with, but I wonder why it would be any different if those few
thousand classes are now packed in a jar and made available as a library to
the webapp?
We run our standalone appserver from a jar, and loading is
Chris,
I don't know what you mean by your response.
-Original Message-
From: Christopher Schultz [mailto:[EMAIL PROTECTED]
Sent: Friday, August 24, 2007 6:43 AM
To: Tomcat Users List
Subject: Re: Need a way to identify tomcat instances at run-time
-BEGIN PGP SIGNED MESSAGE-
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
Gerhardus,
[EMAIL PROTECTED] wrote:
Does that mean that setting settings in
apache:/etc/httpd/conf/workers.properties will override any
workers.properties set in tomcat servers?
Tomcat never deals with workers.properties at all. Only mod_jk
Thanks, I always tough that with ajax requests cookies were not sended.
This is my first application that mixes ajax requests and session
tracking, and becaus I was not able to read session info, I immediately
assumed that (and i think i had read that somewhere). But the problem
was with the
There are some advantages and you can gain higher availability and therefore
improved performance from clustering a single machine.
In my little apps I use terracotta. Open source Java clustering tool that
lets me acheive a pretty good distributed cache without manually rewriting
code to work
It works most of the time but sometimes it doesn't for no discernible reason.
:( One of life's great mysteries that I am looking in to.
In Response To:
Hi Chris,
I guess you're question pointed to Sam, right?
Anyways, that's also what I think.
According to the specs a *noncaught* exception
Hi,
We're hosting a fairly high traffic web application based on Tomcat.
It's running on Debian-Etch, JDK 5.0U10 and Tomcat 5.5.20. We're using
Apache as a front-end with the AJP connector.
The problem I'm encountering is that for a percentage of the POST
requests, Tomcat seems to loose all
I have a struts-based application running on multiple tomcat instances, load
balanced by a hardware load balancer, i.e., no Apache Web Server. I need a
way at run-time to know which tomcat instance it is. Is there a way to
access info in the server.xml or context.xml file at run-time? Can I
Hello,
request.getRemoteUser() returns null in my servlet.
request.getAttribute(REMOTE_USER) also returns null. I have even
checked the headers that are being sent to the AJP connector in Tomcat.
Here's what I'm running: apache-2.0.59 in front of tomcat-5.5.23_1 via
mod_jk-ap2-1.2.23 on
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
Brian,
Brian Barnett wrote:
I have a struts-based application running on multiple tomcat
instances, load balanced by a hardware load balancer, i.e., no Apache
Web Server. I need a way at run-time to know which tomcat instance it
is. Is there a
There may be better ways, but one approach which occurs to me is to set
unique jvmRoute values on the engines in the server.xml. Even though you
aren't using mod_jk, this will result in the provided values being
appended to the sessionid. Assuming that you are establishing sessions,
and that
| From: Brian Barnett [mailto:[EMAIL PROTECTED]
| Sent: Friday, 24 August, 2007 11:54
|
| I have a struts-based application running on multiple tomcat
instances,
| load
| balanced by a hardware load balancer, i.e., no Apache Web Server. I
need a
| way at run-time to know which tomcat instance it
Thanks for the response Christopher... although I had very early on,
already tried what is suggested in the link you have referred to, i.e.,
setting tomcatAuthentication=false in the appropriate server.xml line
(see the config listing I produced earlier in the thread).
Oh I hope its not some
Dan Beaulieu wrote:
Well I guess this gets into the nuances of classloaders that I am completely
unfamiliar with, but I wonder why it would be any different if those few
thousand classes are now packed in a jar and made available as a library to
the webapp?
nothing to do with classloading,
hi,
I encountered a problem with tomcat 5.5. Grazing the whole web i didn't
find any solution (some guys are having the same problem but they also
got no useful hint up to now). The problem is as follows:
I try to build an applet-servlet communication using serialized objects.
In my test
I'm not even concerned with the webapp being automatically reloadable. I can
reload from the manager or just bounce tomcat. Mentioning WatchedResource is
a red herring.
If I war a webapp with many class files in various packages in
WEB-INF/classes with reloadable=false and unpackwars=false it
try adding a larger read buffer
socket.appReadBufSize=34000
or even larger than that.
or apply this patch
http://svn.apache.org/viewvc?view=revrevision=569489, that makes
adjustments for the buffer sizes automatically
Filip
Hoa Doan wrote:
Hi,
I am trying to upload a file size of 2MB to
Omar I actually had this exact same problem early today although I'm
sure my environment is slightly different from your perhaps I can offer
some help.
In my case I have LDAP authentication configured for my servlet. I
believe this step should be the same regardless of the authentication
Dan Beaulieu wrote:
I'm not even concerned with the webapp being automatically reloadable. I can
reload from the manager or just bounce tomcat. Mentioning WatchedResource is
a red herring.
reloadable attribute - nothing to do with reloadable apps
autoDeploy attribute - everything to do with
In addition to my questinos below, is mbeans-descriptors.xml a
Tomcat-specific file or part of JMX in general? I can't tell - it seems to
appear in links like the following:
http://commons.apache.org/modeler/apidocs/org/apache/commons/modeler/package-summary.html
But its DTD is under
Although the debate between session replication vs stateless can't be
definitively solved and depends on the application, I'd like to hear about
your experiences with both. How are you designing your current applications,
and what implications has the choice that you might not have expected -
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
Filip,
Filip Hanik - Dev Lists wrote:
Mentioning WatchedResource is a red herring.
reloadable attribute - nothing to do with reloadable apps
I disagree:
relodable (Context attribute)
Set to true if you want Catalina to monitor classes in
Have you considered distributed caching for session replication? I avoid
stateless like the plague.
There are several tools out there for this, but my favorite (because they
are open source) is Terracotta. It clusters the JVMs.
Native clustering in Tomcat putters out, in my experience, after 3
Hi,
still reading :-) http://tomcat.apache.org/tomcat-6.0-doc/logging.html
I have setup log4j as above, but with a little twist.
If you permission $CATALINA_HOME as read only and you want to customize the
log4j.properties then you need to install the log4j1.2.14.jar and
log4j.properties in
I have a webapp that is working but I wanted to update a servlet jar so:
jar uf MyJar.jar net/whohah/portal/servlet/MyServlet.class
cp MyJar.jar $CATALINA_HOME/webapps/ROOT/WEB-INF/lib
No I get the exception shown below. I have MyServlet stripped down so
the only imports are:
import java.io.*;
Christopher Schultz wrote:
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
Filip,
Filip Hanik - Dev Lists wrote:
Mentioning WatchedResource is a red herring.
reloadable attribute - nothing to do with reloadable apps
I disagree:
relodable (Context attribute)
Set to true if
I'd go the opposite direction, I will do stateless before attempting any
kind of replication or distributed cache.
The scalability of a stateless application will outscale any replicated
app any day, since your fail over options are so much more simplified.
Filip
Dwayne wrote:
Have you
if you want to use log4j in your application only,
then forget everything you read on logging.html
all you need to do is log4j.jar in WEB-INF/lib
log4j.properties in WEB-INF/classes
the steps you are outlining, are converting tomcat from using
java.util.logging to log4j for the container. and
better check your Jar to make sure it has the MyServletException.class
file still in it.
Filip
Dave Sailer wrote:
I have a webapp that is working but I wanted to update a servlet jar so:
jar uf MyJar.jar net/whohah/portal/servlet/MyServlet.class
cp MyJar.jar
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
Lightbulb,
lightbulb432 wrote:
How are you designing your current applications, and what
implications has the choice that you might not have expected - would
you do anything differently if you could?
(For the record, I am not using session
I would like both one common/lib/log4j.properties for the main Tomcat
container logging and another log4j for each web apps - where you set its
log filename and rotation etc...
The tomcat logging would be of interest to an admin and the web app logs to
the developer.
Do I make sense?
Tx -
for tomcat container logging, its easy
1. put log4j.jar in TC_HOME/lib
2. put log4j.properties in TC_HOME/lib
3. compile the extras
ant -f extras.xml
4. replace TC_HOME/bin/tomcat-juli.jar with the one compiled from step 3
5. put extras//tomcat-juli-adapters.jar in TC_HOME/lib
and that's
it does. And if it didn't jar uf is broken, I would think.
On Fri, 2007-08-24 at 15:20 -0600, Filip Hanik - Dev Lists wrote:
better check your Jar to make sure it has the MyServletException.class
file still in it.
Filip
Dave Sailer wrote:
I have a webapp that is working but I wanted to
Hi Robert,
Thanks for the response.
So I've come to believe that its possible to avoid using Tomcat
authentication altogether, i.e., without specifying realms and using
tomcat user/roles in an application's web.xml. Given my context (a
University environment with over several hundreds of
Hi Robert,
Thanks for the response.
So I've come to believe that its possible to avoid using Tomcat
authentication altogether, i.e., without specifying realms and using
tomcat user/roles in an application's web.xml. Given my context (a
University environment with over several hundreds of
73 matches
Mail list logo