Hi all,
We have a configuration where we use the tomcat connector with the Sun
Java System Web Server and I'm noticing an unusual behavior that I've
not noticed with Apache and mod_jk.
When making a request to a URL such as
http://host/webapp/dir/
In apache, the underlying request is made as:
RE
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
Wolfgang,
(I'm cc'ing the tomcat-user mailing list in case others would like to
read my response. Please try to keep things on the list rather than
emailing contributors directly.)
On 10/5/2010 9:34 AM, Wolfgang Orthuber wrote:
> thanks for your deta
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
Rob,
On 10/5/2010 11:28 AM, Rob Gregory wrote:
> Basically we are stuck with some legacy application parts that while
> these are scheduled to be replaced we have to support them until they
> have been. Using filters would not solve the issue as the '
Hi Maximilian,
Basically we are stuck with some legacy application parts that while
these are scheduled to be replaced we have to support them until they
have been. Using filters would not solve the issue as the 'Hack' as you
put it is done at a cookie path level. We do use filters to implement
se
This application (or mess of applications munged together) is a big mistake,
the source of all your problems and should be the only thing that you address.
That's why nobody else has "this problem". "This problem" is a sign of serious
misconceptions in design and development.
And if you really
Hi Chris,
In Internet Explorer 5.5,6.0 if you opened up two separate browser instances
they would have two 'un-connected' sessions. As stated by Ronald they would
share the session if the 2nd was opened using ctrl-n but otherwise the sessions
would be unique. Cookies may have always worked as t
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
Rob,
On 10/5/2010 4:26 AM, Rob Gregory wrote:
> Is there any way to dynamically create these contexts or do they
> require a live.xml, test.xml, etc within conf/Catalina/localhost.
You can use JMX or the manager to deploy an application under any
con
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
Rob,
On 10/5/2010 5:03 AM, Rob Gregory wrote:
>> Sounds like you need to be pretty careful. Is it possible you've built a
>> fragile application?
>
> Some legacy parts of the application became fragile when the browsers
> started sharing sessions and
Hi Jeffrey,
Yes, the cookie can and is tied to the context. I just take this another
level and tied it against a virtual context so as far as the browser is
concerned they are different sites and as such, different sessions.
Regards,
Rob.
> -Original Message-
> From: Jeffrey Janner [mai
I agree, this is a very scary app to support.
At a minimum, multiple contexts, each with its own database resource
definition would avoid a lot of the OP's problems, wouldn't it? I'm not
a developer, but from an SA standpoint, this gives me the willys.
Q: Can the cookie be tied to "hostname/conte
Our app running on Tomcat 5.0.28 has some kind of kluge involving a
flown-in vhosts.xml file in .../conf that uses aliases such that people
can have separate sessions of the same app on separate tabs. I have no
idea if this is how this should have been done back in the olden days
but more impo
On 05.10.2010 15:27, mamalacation wrote:
n828cl,
n828cl wrote:
Pretty much guaranteed that it's not Tomcat but your webapp that is
locking itself out of access to some resource (such as the database).
Take several thread dumps during the slowdown period and see what's going
on.
http://wiki
Hi Chuck,
It wouldn't be live test and train, normally it's test & train together
with live being a separate install altogether. Unfortunately, this is a
requirement I cannot avoid and was fine until the browsers changed to
share sessions. I presume they have changed to support navigating the
same
Op dinsdag, 5 oktober 2010 15:27 schreef mamalacation
:
n828cl,
n828cl wrote:
>
>
> Pretty much guaranteed that it's not Tomcat but your webapp that is
> locking itself out of access to some resource (such as the database).
> Take several thread dumps during the slowdown period and
> From: Rob Gregory [mailto:rob.greg...@ibsolutions.com]
> Subject: RE: Tabbed browsers sharing session - work around.
> I am surprised this hasn't been a problem for a
> lot more people.
It's not a problem for most because most don't try to run live, test, and
training inside a single on a s
n828cl,
n828cl wrote:
>
>
> Pretty much guaranteed that it's not Tomcat but your webapp that is
> locking itself out of access to some resource (such as the database).
> Take several thread dumps during the slowdown period and see what's going
> on.
>
> http://wiki.apache.org/tomcat/HowTo#Ho
Hi Stephen,
The approach we use was described in my original post. If you need more
information then I can explain further if it helps. Note, our approach
does require a bespoke tomcat request class (to change the session
cookie path) and this has not been without teething problems. Are you
also u
Rob,
The way you describe session sharing is indeed a problem. The way we deal with
this is to use a separate database table to keep track of window ids. A unique
value is assigned when a window is opened and maintained until the window is
closed. Although the session may be the same for all
Hi Ronald,
Yes I was aware of that behaviour... Just for reference Firefox and
Google Chrome also share session logic so I am surprised this hasn't
been a problem for a lot more people. I am happy that my work around
solves the session sharing problem but would still prefer to go down the
dynamic
Rob,
IE 6 is even more confusing. If you open a new window with ctrl-N you have the
same session sharing as with tabs. Only if you click the IE6-icon to start a
new instance of the process it will not share them. Opening a new tab in IE7 is
like using ctrl-n to open a new window in IE6.
Ronal
> From: mamalacation [mailto:mamalacat...@hotmail.com]
> Subject: tomcat 6.0 bottleneck
> From what I understand is that my system has still adequate free physical
> resources, sql server is not being pushed and the bottleneck comes from
> tomcat, where I have no clue what to tune.
Pretty much g
2010/10/5 Juan Carlos Afonso :
> Hello,
>
> My tomcat6 executes as a daemon and I need to know where is its home
> directory, because tomcat6 user executes a pg_dump command and it needs to
> find the .pgpass file. Thanks in advance.
>
You mean System.getProperty("user.home"); ?
http://downloa
...and another thing:
Yesterday, the server was running under this load, so when I checked the
manager/status today I saw that current thread count was 500 while busy
threads were only about 10. Shouldn't the threads be killed at some moment?
What is the timeout option for this behavior?
Thanx a
Dear all,
I administer a tomcat 6.0 on a windows 2008r2 server with sql server 2008. I
am a Unix admin that had no contact with tomcat in the past; so, both tomcat
and windows are a very peculiar combination for me :)... This server runs an
application that can become quite (but-not-that-much) lo
Chris,
See comments below:-
> Rob,
>
> On 10/4/2010 7:27 AM, Rob Gregory wrote:
> > Using the hostname doesn't really guarantee a unique session for example
> > if I click new tab and paste the URL into the new window I suspect the
> > browser will see the same session from the first tab.
>
> No
Hello,
My tomcat6 executes as a daemon and I need to know where is its home directory,
because tomcat6 user executes a pg_dump command and it needs to find the
.pgpass file. Thanks in advance.
Hi Chris,
Is there any way to dynamically create these contexts or do they require a
live.xml, test.xml, etc within conf/Catalina/localhost. The multiple contexts
would be my preferred approach although I would like to achieve this with a
single code base if this is possible. The multiple envir
27 matches
Mail list logo