In OSX you must use ipfw, i don't know how ipfw is used, but i think this
must be similar to iptables.
On 8/15/07, Stephen Caine <[EMAIL PROTECTED]> wrote:
>
> David,
>
> Do you know what the corresponding command (iptables) would be in OS X?
>
> > You can execute tihs iptables line (if you are us
Hi all,
after moving a jar file from tomcat/common/lib to tomcat/shared/lib, JSPs
that use that library don't compile anymore, it says package does not
exist. It is a library from my company, used by all webapps in the server
installation. It used to be in tomcat/common/lib as it was used by a
to
Tony Fountain wrote:
Hi,
Scenario: we purchased a product written in Java to integrate into our
reporting tool. Our setup is such that our web application is written
in .NET and hosted on a web farm using IIS (5 or 6 depending on the
environment). The product we purchased runs under Apache Tom
Hi Ole,
when you tried against your installed APR most likely the dev rpm, which
includes the necessary header files for compiltions where not installed.
When you tried against a BUILD directory, most likely the APR library
need for linking was not there.
Whichever way you choose, you need
nageshsrao wrote:
we are getting the GC printed on to the same catalina.out and we see that the
memberAdded messages appear almost at the time of GB getting printed, does
it prove that longer GC pauses are causing this? is there any other data
points/proof can be get?
E.g. -XX:+PrintGCApplicati
As you guessed would be intuitive, yes, the JSP compiler include classes
in shared/lib (and also classes in webapp itself).
The shared/lib is the same as if you copy it's content to the
WEB-INF/lib folder. As your compiler does not include the shared/lib in
it's classpath, there can be various reas
Someone have a case study wich compare Tomcat with others application
servers?
Thanks a lot
Andrew
I assume you've already Googled what you're looking for and not found
anything? What are you after exactly, performance or feature comparison?
please be more specific...
I assume you're after more than this:
http://en.wikipedia.org/wiki/Comparison_of_application_servers
Andrew Hole wrote:
Som
(see below for message context)
Ok, I've decided on using Http Basic authentication for my web service,
and successfully configured tomcat to authenticate against the
tomcat-users.xml file to the point where I can access a valid principal.
But now I've got another problem.. :-)
I tried acces
oops, also here is my resource definition from my web.xml:
Link to the UserDatabase instance from which we request lists of
defined role names. Typically, this will be connected to the global
user database with a ResourceLink element in server.xml or the context
confi
InitialContext.lookup() gives you a simple object:
so change your code to
Context ic = new InitialContext();
Object o = ic.lookup("java:comp/env/users");
set a breakpoint and see, what type of object you're getting back.
hth
gregor
--
what's puzzlin' you, is the nature of my game
gpgp-fp: 79A
Hi David,
thanks for your answer. What I forgot to say is that I use Tomcat 5.0.28.
I'll comment inline to your options.
On Thursday 16 August 2007 10:41:30 David Delbecq wrote:
> As you guessed would be intuitive, yes, the JSP compiler include
> classes in shared/lib (and also classes in webapp
Hi,
I want to track all the active sessions of a tomcat 5.5 cluster.
Particularly, I want to host a page on each cluster member which will
display all the sessions active on its cluster.
So I want to see a list of :
session-id, user-name, jvmRoute (of member to which this session is sticky
to).
Try this then - this is my standard character encoding index.jsp test.
<%@ page contentType="text/html; charset=UTF-8" %>
Character encoding test page
Data posted to this form was:
<%
request.setCharacterEncoding("UTF-8");
out.print(request.getParameter("
//code
Object o = ic.lookup("java:comp/env/users");
System.out.println(o.getClass().getName()); // prints :
"org.apache.catalina.users.MemoryUserDatabase"
doing instanceof tests on the returned object for MemoryUserDatabase &
UserDatabase all fail, even though in debug that's clearly what it
this is weird, check this out:
//code (tomcat 5.5.23)
java.security.Principal p = request.getUserPrincipal();
System.out.println(p.getClass().getName().equals(MemoryUser.class.getName()));
// prints "true"
System.out.println(p.getClass().equals(MemoryUser.class)); //prints "false"
So what this
Hi
I don't think webalizer is 'for' apache httpd, rather it will analyse
logs written in Common Logfile Format (and others). It doesn't matter
what he server is.
If you use the AccessLogValve to write your access logs then webalizer
seems to read these well. I'm a raw beginner with webalizer. I ma
I'm upgrading from tomcat 4.1.24 to tomcat 5.5.12 on WinXP. I have three
applications that are deployed from this
server, two of which work fine with tomcat 5.5.12, however the third which uses
an oracle 9i database gives me the
following error "javax.servlet.ServletException: Cannot create JD
> From: Matthew Kerle [mailto:[EMAIL PROTECTED]
> So what this is saying is that the *names* of the classes are
> the same,
> but the actual classes are different. this is crazy...
I suspect the two classes are being loaded by different classloaders - a
common and entertaining* problem in Tomca
hmm, you have an interesting problem!
first thing I'd say is use this opportunity to upgrade to the
ojdbc14.jar, which is the latest oracle jdbc driver and allows lots of
nice enhancements.
second, I'm assuming that since the error is a servlet exception, that
there's some servlet code manua
Peter,
you're exactly right.
***code***
Class c1 = request.getUserPrincipal().getClass(); //get the class of the
Principal that tomcat created , which is a MemoryUser instance
Class c2 = MemoryUser.class; // get the class loaded by the current loader
System.out.println(c1.getClassLoader().getCl
check this
http://evolutionnext.com/blog/2005/10/13/1129259088959.html
On 8/16/07, Matthew Kerle <[EMAIL PROTECTED]> wrote:
>
> hmm, you have an interesting problem!
>
> first thing I'd say is use this opportunity to upgrade to the
> ojdbc14.jar, which is the latest oracle jdbc driver and allow
My laptop is having very low memory. I need to run minimal version of
tomcat. (stripped version of tomcat, to avoid memory hogging). Can you guys
give some tips on this -- like removing unnecessary jars & enabling dynamic
class/jar loading?
(My requirement is just to run few jsp pages, to enter/re
> From: Matthew Kerle [mailto:[EMAIL PROTECTED]
> Class c1 = request.getUserPrincipal().getClass(); //get the
> class of the
> Principal that tomcat created , which is a MemoryUser instance
> Class c2 = MemoryUser.class; // get the class loaded by the
> current loader
> System.out.println(c1.ge
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
Leon,
Leon Rosenberg wrote:
> security by obscurity, that is.
Through I agree that claiming Apache httpd increases security, it is
certainly not security by obscurity. It is another layer between the
attacker and the goodies. You may disagree about t
taking a look at tomcat 5.5.23, i'll try to answer you
1) you can take a look at tomcat embedded, it *may* be stripped down,
but am not sure
2) in the core tomcat, there are perhaps only 2 or 3 jars you can
*perhaps* remove: tomcat-18n other than english, tomcat-ajp,
catalina-optional
3) as for yo
Best Regards,
Apoorv
> _
> From: Garg, Apoorv
> Sent: Thursday, August 16, 2007 9:14 AM
> To: 'users@tomcat.apache.org'
> Subject: Tomcat freezes up
>
> Hi,
>
> I am using Tomcat 5.5.17. I have noticed that at times Tomcat freezes
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
DAvid,
David Hesson wrote:
> The
> content-length has a maximum value of 2.x billion, which is right under
> two gigabytes.
Is this a limit on commons-upload? The header itself can contain an
arbitrarily high number, so there's no inherent file-size
you're exactly right again. I just checked my project settings, I had to
add catalina.jar to the project libraries to get the class to compile,
but I'd forgotten to prevent it from being deployed, so there was a copy
of catalina.jar in my /WEB-INF/lib, doh!
So I configured it to not be deploye
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
Matt,
Matthew Kerle wrote:
> this is weird, check this out:
>
> //code (tomcat 5.5.23)
> java.security.Principal p = request.getUserPrincipal();
> System.out.println(p.getClass().getName().equals(MemoryUser.class.getName()));
> // prints "true"
> Sys
> From: Matthew Kerle [mailto:[EMAIL PROTECTED]
> the MemoryUser class is in catalina.jar, which is in the server/lib
> folder. would I be right in saying that web application code
> is barred
> from loading any classes from the server/lib directory?
(light bulb comes on)
Ah yes, I remember t
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
Peter,
Peter Crowther wrote:
> We ended up with the horrible, horrible hack of
> pulling the class out of catalina.jar, putting it in its own jar, and
> deploying that in common/lib.
Shouldn't it be acceptable to simply move catalina.jar from server/
Hi Chris
I naively tried relocating the catalina.jar to /common/lib, and got the
below error. Peter has a good comment to this problem in his reply, so
I'll continue the thread in response to his mail.
many thanks!
cmd /c C:\servers\apache-tomcat-5.5.23\bin\catalina.bat run
Using CATALINA_BA
> From: Christopher Schultz [mailto:[EMAIL PROTECTED]
> Shouldn't it be acceptable to simply move catalina.jar from server/lib
> to common/lib?
>
> Sure, you'll still have a non-standard install, but it's easier to
> script a setup like that than pulling specific classes out of
> the distro (whi
no, see my previous reply, tomcat fails to bootstrap if catalina.jar is
not in server/lib...
Christopher Schultz wrote:
Peter,
Shouldn't it be acceptable to simply move catalina.jar from server/lib
to common/lib?
Sure, you'll still have a non-standard install, but it's easier to
script a setu
I agree, the Principal interface is verily hobbled and almost useless
(Go Sun!). The catalina implementations are much more user-friendly, but
unfortunately difficult to access.
I can't really justify making the tomcat install non-standard (also
probably not possible as it's owned by the clien
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
Matthew,
Matthew Kerle wrote:
> no, see my previous reply, tomcat fails to bootstrap if catalina.jar is
> not in server/lib...
That's too bad.
Why not just use the built-in authentication and authorization mechanism
instead of trying to use Tomcat's
just downloaded security filter and had a look, it looks very cool. If I
had more robust requirements for my authentication (and more time!) I
would probably use it.
At the moment though I've got a workable work-around in using the
toString() method, so I'll just use that instead.
thanks Chr
> From: Matthew Kerle [mailto:[EMAIL PROTECTED]
> Subject: Re: catalina error log
>
> you're missing the Apache Portable Runtime library from your PATH
> variable. this is not a serious problem, but if you don't
> want to get this error then download the version of the library
If you don't ne
Hi!
> A more flexible option is to use securityfilter
> (http://securityfilter.sourceforge.net) to handle everything.
>
If you are already using spring have a look at ACEGI.
It is not really easy to install, but allows you to e.g. have different
login methods within the same webapp.
Regarding t
I posted the same question week back. People advised me to handle the
resource safely, like freeing up the database connections after use,
releasing file handles etc. I am very sure that I am doing that perfectly,
but still my tomcat freezes over long run. I've seen this case in many
instances. Is
http://tomcat.apache.org/tomcat-5.5-doc/class-loader-howto.html
this is why I can't reference any classes loaded from server/lib in my
webapp, the server/lib classes are loaded by the web application
classloader's "uncle", so to speak, the sibling of it's parent. so it
makes sense that no web
Tomcat 5.5 has an all-to-all replication mechanism, hence all nodes
should be identical. you would only need to see the sessions on one node
to get a picture of what is looks like elsewhere.
there is a flag called "notifyListenersOnReplication" this is only for
attribute replication for nodes
It seems very unlikely to me that the problem is in the Tomcat code. It
is very widely deployed and any deadlocks would be found relatively
quickly unless you use a very obscure setup.
Did you do any standard deadlock debugging? E.g., thread dumps and
deadlock analysis when the freeze occurs,
Hi Jung,
That did the trick!
I downloaded and installed:
http://www.axint.net/apache/apr/binaries/rpm/i386/apr-devel-1.2.8-1.i386.rpm
Then I ran this:
./configure --with-apr=/usr/bin/apr-1-config && make && make install
That did the trick :-)
Thanks Jung, Hassan, and Stephen,
- Ole
Rainer
Mario, you are a hero. do women come and worship you in the street? they
should! Using reflection to break into an object of a foreign class is
just...genius! this is the sort of thing that Ruby programmers do all
the time, but is very hard to do in Java...
my final code (in the context of a S
Hi,
When I start tomcat (5.5.16) as a non-root user (tomcat) I am unable to
access anything I have set up in my context container and I see nothing
in the logs to help (the access log says 404 for all the requests). For
example, I have the standard setup:
/usr/local/tomcat/latest/conf/Catalina/lo
There were in fact bugs with requests > 2GB in both Tomcat and mod_jk
which were fixed recently, as Rainer pointed out. (And some of the
code was using -1 if the content-length was too large.) Upgrading to
6.0.14 or the next 5.5 release should fix the problem.
There's no size limit on commons-file
Hi!
> Mario, you are a hero. do women come and worship you in the street?
Haha! Oh boy ... you don't want to know ...
Glad it helped you! :-)
Ciao,
Mario
-
To start a new topic, e-mail: users@tomcat.apache.org
To unsubscribe, e
Thanks for your advice so far
I've upgraded to the latest Oracle ojdbc14.jar and placed it the
myapp\WEB-INF\lib folder, I also tried it the
Tomcat\common\lib for for good measure but still got the same results. I'll
continue to use ojdbc14.jar from now
though.
I've change the ResourceParams
why don't you use this jar? ojdbc14_g.jar
That's what I ended up upgrading to.
Let me know if you need it, Ian, and I'll send it to you offline.
Barry
And, you should put it in the common/lib folder.
-Original Message-
From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED]
Sent: Thursday,
Hi Ian.
ojdbc14.jar needs to ONLY be in common/lib. It won't work in
myapp/WEB-INF/lib and won't work if you have the jar in both places.
Additionally you should take a look at the JDBC howto docs regarding
what your config in your myapp/META-INF/context.xml or
conf/Catalina/localhost/myap
Hi David
David Smith wrote:
My only editorial comment on the page is to NOT place your
or definition in server.xml as
recommended on the page. Place it in context.xml or myapp.xml as I
describe above.
Just quickly, I was wondering why you recommend this? I know the tomcat
docs have chang
looks like your xml doc has an un-closed tag or similar, hate to suggest
this but maybe can you recheck your change to make sure this isn't the case?
I'd suggest going with David's suggestion, and put your
definition in a separate file called 'context.xml'. Explode your WAR,
put this in /META
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
Matt,
Matthew Kerle wrote:
> I never
> understood why, and personally doing things that way is a serious pain
> for me since it means I need to build a separate deployment descriptor
> for dev, test & prod, which means I need to know the prod database
In my experience, a resource is usually only relevant to one webapp.
There's no need to put it in server.xml as a GlobalNamingResource unless
you want that resource available in all your webapps. Moving the
resource to the block of a context.xml file also makes it so
resources can come and g
On 8/16/07, Len Popp <[EMAIL PROTECTED]> wrote:
> You're right that putting server-specific info in context.xml in the
> .war is no good when the app could be installed on different servers.
Isn't that what build.properties files are for? :-)
--
Hassan Schroeder [EMAIL
Hi Chris
my experience has been: one tomcat instance per dev/test/prod, & the
webapp on each instance always points to the same database (dev
tomcat->dev db, test tomcat -> test db etc), so the database connection
on a /single/ instance never changes. However when promoting code from
dev to t
Hi everyone.
I'm testing this code:
${ table_data.package_price +
( road_assistance_2year == '-' ? 0.0 : road_assistance_2year ) +
( comp_offer_data.hasInsurance ? table_data.insuranceValue : 0.0 ) }
and in Tomcat 6.0.13 it works fine.
However in Tomcat 5.0.19 the page stops proccessing a
How do you use a build.properties file to define a JNDI resource? Or
do you create JDBC connections a different way?
--
Len
On 8/16/07, Hassan Schroeder <[EMAIL PROTECTED]> wrote:
> On 8/16/07, Len Popp <[EMAIL PROTECTED]> wrote:
>
> > You're right that putting server-specific info in context.xml
The reason is that server.xml is for the whole server, so when you
change something you have to restart the whole server rather than just
the one application.
You're right that putting server-specific info in context.xml in the
.war is no good when the app could be installed on different servers.
now that sounds good! the only thing is I don't see how that maps to a
DataSource declaration, the element in
doesn't seem to allow the full range of
properties that you need to define a database connection, eg -
username/password/driverClassName/url etc...
Where would you define these?
Da
On 8/6/07, Dimitris Mouchritsas <[EMAIL PROTECTED]> wrote:
>
>
>
>
>
>
>
>
Maybe just something like:
? HTH,
--
Hassan Schroeder [EMAIL PROTECTED]
-
To start a new topic, e-ma
On 8/16/07, Len Popp <[EMAIL PROTECTED]> wrote:
> How do you use a build.properties file to define a JNDI resource? Or
> do you create JDBC connections a different way?
Not sure I understand the question -- we're talking about putting
a Resource element in META-INF/context.xml, right? So the detai
I don't get it. Are you assuming that the programmer would put the
JDBC server info in a build.properties file that is used at compile
time? That doesn't work unless the programmer knows all the servers
the app will eventually be deployed on - which isn't always the case
for me.
--
Len
On 8/16/07
That was the next thing I looked into after I solved this bug. I
"fixed" it by commenting out the line in
the server.xml file.
My take on the native APR stuff (which is likely completely wrong
since it consists of what I learned in about a 2 minute skim of some
documentation) was that it was rea
it's doable, some of my teammates use that method, but it's a pain in
the butt because you basically have to create 3 WARs per release (one
for each target). considering a struts/hibernate/spring/xFire
application is about 30mb in 3rd party jars, that's a lot of space. ok
HD space is cheap, but
no worries. I hadn't seen how to remove that dependency before, which is
why I didn't mention it in my reply. the APR just gives tomcat
comparable file reading performance as httpd, but unless you're running
slashdot or hea.net on your tomcat then not having it isn't a big issue
and your distr
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
Matt,
Matthew Kerle wrote:
> Most of the sysadmin's I've worked with really don't like the idea of
> having to rip open a war and update some weird config file just to do a
> deploy. the alternative is I know all the passwords and build a war for
> ea
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
Len,
Len Popp wrote:
> How do you use a build.properties file to define a JNDI resource? Or
> do you create JDBC connections a different way?
Search-and-replace, baby. That's how I have my configuration working.
- -chris
-BEGIN PGP SIGNATURE
Compare performance...
On 8/16/07, Matthew Kerle <[EMAIL PROTECTED]> wrote:
>
> I assume you've already Googled what you're looking for and not found
> anything? What are you after exactly, performance or feature comparison?
> please be more specific...
>
> I assume you're after more than this:
>
Hi Rainer,
Thanks again for that great fix. When I fired up Tomcat, I still get this
message:
Aug 16, 2007 9:53:05 AM org.apache.catalina.core.AprLifecycleListener init
INFO: The Apache Tomcat Native library which allows optimal performance in
production environments was not found on the java.
On 8/16/07, Christopher Schultz <[EMAIL PROTECTED]> wrote:
> They're all crap solutions when nobody wants to do their job.
heh -- indeed.
And how about we add the "it depends" corollary -- they're all
crap solutions for /some/ specific environment(s) :-)
--
Hassan Schroeder --
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
Matt,
I disagree with many of your assertions.
Matthew Kerle wrote:
> it's doable, some of my teammates use that method, but it's a pain in
> the butt because you basically have to create 3 WARs per release (one
> for each target). considering a stru
On 8/16/07, Len Popp <[EMAIL PROTECTED]> wrote:
> I don't get it. Are you assuming that the programmer would put the
> JDBC server info in a build.properties file that is used at compile
> time? That doesn't work unless the programmer knows all the servers
> the app will eventually be deployed on -
I am testing our webapp under tomcat 6 after being under tomcat 5.5.
I have moved the following file from 5.5. into 6. I have placed it in the
conf directory:
This works under 5.5, but under 6.0, I get "Cannot create JDBC driver of
class '' for connect URL 'null'
And
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
Len,
Len Popp wrote:
> I don't get it. Are you assuming that the programmer would put the
> JDBC server info in a build.properties file that is used at compile
> time?
Well, you have to get creative. For instance, all our devs (and
deployment folks)
As far as I'm aware, there is no difference between a
element in context.xml and a element in a
... block. Well... other
than the need to use a to make it available to an
individual webapp. Did you try it and get a failure?
--David
Matthew Kerle wrote:
now that sounds good! the only t
Hi Chris
I apologise for coming across a bit harsh there, don't mean to offend!
I'll take it a bit easier...;-p
Either of the methods you recommended looks good, the only trouble is
that the doco is a bit... terse on the subject of setting up a JNDI
datasource outside the war file, and it's
Hmmm. I'm thinking I'm beginning to get a whiff of the issue.
Before, we would run multiple webapps and each had their own context file
that matched the webapps. Like for the one below, we had one called mo.xml.
We'd put all these in the CATALINA_HOME/conf/Catalina/localhost and it all
worked grea
Hmm. Sounds like, yeah, this is what I need to do too.
-Original Message-
From: Matthew Kerle [mailto:[EMAIL PROTECTED]
Sent: Thursday, August 16, 2007 1:37 PM
To: Tomcat Users List
Subject: Re: JDBC driver of class '' for connect URL 'null' in Tomcat 5.5.12
Hi Chris
I apologise for com
no I didn't...:-(
I was going off the globalresources config reference page, wher it lists
the attributes of a element as being:
**snip**
The valid attriutes for a element are as follows:
Attribute Description
auth
Specify whether the web Application code signs on to the corresponding
I just upgraded from 5.5 to 6.0 without any problems. My webapp was
deployed as a war file
with the config.xml in the META-INF (the exact config for 5.5). The
deployment does the
correct thing with the config.xml.
Make sure your jdbc library is in the CATALINA_HOME/lib dir.
I also was playing
I'm not sure about Tomcat 6, but in Tomcat 5.x there's no such thing
as $CATALIN_HOME/lib. YOur JDBC-drivers should either be bundled with
your web-app (if that's the only one using them) or, if used both from
Tomcat and your webapp(s) to $CATALINA_HOME/common/lib
Coming to the context:
- Create
I see. Some of the tomcat docs are probably out of date, depending on
someone pointing out the flaw or contributing updates.
Tomcat uses reflection to configure the pool using bean setter methods.
That makes the full set of attributes dependant on the implementation
used -- in this case DBCP
Mark Thomas wrote:
request.setCharacterEncoding("UTF-8");
Is this always safe? For responses I can (and do) check the
accept-charset request paramater, but I can't figure out how to tell
what the request encoding should be.
--
Right ... tomcat 6's version of common/lib/ is simply lib/
Location of the driver jars is not entirely as simple as either in your
webapp or on the container's lib directory. Like the Highlander (good
movie if you haven't seen it), there can be only one. If another webapp
is using it and it'
Hi Chris
It seems like you guys have a pretty good setup going, better than the
one at my (current) company anyway. I didn't realise that your
environment doesn't have dependencies on statically linked war files
like ours does. The developer makes a release and creates the three wars
for depl
much better, thank you! I think I'm gonna have to put aside some time
to brush up on tomcat 5.5/6 jndi datasourcing as I'm obviously out of date.
cheers!
David Smith wrote:
I see. Some of the tomcat docs are probably out of date, depending on
someone pointing out the flaw or contributing upd
OK, I was trying to go away from this approach because Tomcat 6 didn't ship
with a conf/Catalina/localhost directory, but apparently, this is still the
way to do things. So I created this directory, put the xml files that synced
with my webapp in there and everything's back to working peachy keen.
I see, so things quite changed from 5 to 6, however:
When it comes to a Tomcat 5.x, it still goes that you put your jar
(i.e. a jdbc-driver)
EITHER
in $CATALINA_HOME/lib (when used from Tomcat AND other webapps)
OR
in $CATALINA_HOME/webapps/yourwebapp/WEB-INF/lib (when used by a
specific web-a
Where do you put tcnative.so?
And: if you do ldd PATH_TO_TCNATIVE/tcnative.so: are there any
dependencies shown, which do not lie in /lib or /usr/lib, or which ldd
can not resolve? If yes: which libraries, and which path resp. which
libraries without path? Maybe just post the result of the ldd
Hi!
How can I solve the problem to point JNDI-Resources to different values
for the same webapplication that is deployed in serveral hosts on a
server without having to rewrite the context.xml in the webapp for each
deployment?
with using jndi-resource points to
resource1 defined in Glob
I just checked the docs and the order in which classloaders are searched
hasn't changed from 5.5 -> 6.0. Still have to be careful a specific
class is only found once in the classloader tree from the perspective of
the webapp.
--David
Gregor Schneider wrote:
I see, so things quite changed f
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
Matt,
Matthew Kerle wrote:
> the doco is a bit... terse on the subject of setting up a JNDI
> datasource outside the war file, and it's not as straightforward or easy
> to find as compared to the method for placing the context.xml inside the
> deploym
A comet read event doesn't update the last accessed time of an
HttpSession -- which means comet read events will never prevent an
http session (not the comet session) from timing out and the
connection getting closed.
Is that by design? If so, can anyone offer me suggestions on how to
handle long
Peter Warren wrote:
A comet read event doesn't update the last accessed time of an
HttpSession -- which means comet read events will never prevent an
http session (not the comet session) from timing out and the
connection getting closed.
Is that by design? If so, can anyone offer me suggestions
Rainer Jung wrote:
Where do you put tcnative.so?
more specifically, is LD_LIBRARY_PATH pointing to the directory of the
.so file?
Filip
And: if you do ldd PATH_TO_TCNATIVE/tcnative.so: are there any
dependencies shown, which do not lie in /lib or /usr/lib, or which ldd
can not resolve? If
This is more of a Servlet API question, but pertains to Tomcat
nonetheless. I am using Tomcat 5.5.23 with JDK 1.6.
I am having a problem with Generics and request.getParameterNames().
I keep getting a unchecked cast warning (just a warning, but still).
I was going to just use the @SuppressWarnin
Hi Rainer,
Thanks for your suggestion! Using the specs file as suggested below did
not work.
Also, gcc -dumpspecs > /path/myspecs and then changing the *libgcc rule
did not
work either.
What finally worked was
export CC="gcc -static-libgcc"
But then I had to recompile OpenSSL, APR and then
1 - 100 of 128 matches
Mail list logo
