irclog2html for #tomcat on 20040408

00:09.32*** join/#tomcat vinse_ (~delfonic_@66-151-49-205.sog.polarisnet.net)
01:12.28*** join/#tomcat ps` (~ps@yoda.rsn.bth.se)
02:04.46jasonbibot: tomcat-benchmarks is http://www.mail-archive.com/tomcat-dev@jakarta.apache.org/msg56210.html http://www.webperformanceinc.com/library/ServletReport/
02:04.47ibotjasonb: okay
02:04.47jasonbibot: wake up!
02:04.49ibotup!: GOOD MORNING!!!
02:04.54jasonbheh
02:11.29vinse_hi jasonb
02:13.45jasonbhi vinse_.
02:14.09jasonbibot: php-from-tomcat is http://wiki.apache.org/jakarta-tomcat/UsingPhp
02:14.10ibotjasonb: okay
02:15.20vinse_jasonb
02:15.39vinse_i'm not understanding how container managed security is saving sesssion state
02:15.42vinse_where are my cookies?
02:18.30jasonbWell, with the exception of the session cookie itself (JSESSIONID), all other cookies are request/response-specific.. so they're stored in the request and response objects.
02:19.01jasonbThe JSESSIONID cookie is really a session ID that is stored in the session object.
02:19.21vinse_i.e. session.getID() ?
02:19.32vinse_where is it stored on the client side?
02:19.51jasonbBut, when a session-active request comes in, the browser sends the JSESSIONID cookie to Tomcat, and Tomcat puts that into the request object.  I think it automatically also sends it out with the response.
02:20.07vinse_well
02:20.13jasonbYeah, Session.getID() or something like that.
02:20.32jasonbOn the client side, cookies are stored however the http client wants to store them.
02:20.34vinse_when i call request.getCookies() i get an array of size zero
02:20.54vinse_although the api sez i should get null
02:20.57vinse_if there are no cookies
02:21.29jasonbOkay, so if your request has no cookies that means that 1) the http client sent no cookies with the request, and 2) Tomcat doesn't know of a session ID cookie for this request (not a logged-in user?)
02:21.39vinse_no it's a logged in user
02:21.58vinse_becuase several isUserInRole calls return the correct result
02:22.31jasonbHmm.
02:22.42vinse_that's what i thought was strange
02:22.47jasonbYou're right about the getCookies() returning a null if there aren't any.. that's what the servlet API says.
02:22.51jasonbWhich version of Tomcat is this?
02:22.59vinse_most recent
02:23.01vinse_5.0.x
02:23.20vinse_but if there's a logged-in user, why would there not be any?
02:23.29vinse_when there is no logged in user it returns null
02:23.40vinse_when there is a logged in user it returns an empty array
02:24.34vinse_i think
02:24.38jasonbHmm.. that could actually be a TC5 bug.
02:25.08vinse_maybe JSESSIONID is a "virtual" cookie, but doesnt end up in the Cookies[] array?
02:25.25jasonbI haven't tried the same thing between TC4 and TC5, but if that's the behavior.. and if it's repeatable, you should probably file it as a bug.
02:25.32vinse_ack
02:25.41vinse_filing it as a bug means i'm not fixing it =O
02:25.43jasonbNo, JSESSIONID is really a cookie when there is one, and it should be in that array.
02:25.51vinse_i guess it's not hurting me any
02:26.00vinse_i can still invalidate the session when they log off
02:26.01jasonbvinse_: If you file that as a bug, it's possible they'd fix it fast.
02:26.22vinse_ok, i'll make sure that's what's happening and file it
02:26.53jasonbYes, please do.
02:27.09jasonbThat is a good find, if it's true, and that could be causing others trouble.
02:27.26vinse_should i be keeping my authorization when i close the browser while using form authentication and container managed security?
02:27.27jasonbTC5 is stable, and has been tested quite a bit, but things like this could fall through the cracks.
02:27.29vinse_because i'm not
02:28.16jasonbThat depends on what kind of cookie JSESSIONID is.. there are two kinds: one that is stored in RAM and goes away when you restart your browser, and another kind that actually persists to disk.
02:28.20vinse_i have to relog in regarless of the session timeout value when i open a new browser, if that's not supposed to happen i'd guess it's a symptom of this bug
02:28.33jasonbI'm not remembering how to choose which type of cookie it is, though.
02:28.43vinse_yeah i couldnt find that anywhere
02:28.59vinse_but from what i read it seemed like a persistent cookie was the default
02:29.03vinse_in the realm how-to
02:33.45vinse_i did that too
02:33.48vinse_saw none from localhost
02:35.00vinse_ima go home now and play w/ it tonight, i'll let you knwo what happens
02:36.04jasonbOkay.. thanks!
02:36.21vinse_thank you ;)
02:36.22vinse_again
02:39.04*** join/#tomcat matroska (~chatzilla@c-67-169-52-64.client.comcast.net)
02:40.07matroskai am getting this erroe : what doe sit mean -     * Deploy error : uPortal : org.objectweb.jonas.web.JWebContainerServiceException : Cannot read the deployment descriptors '/E:/subversion/jonas-3-3-1/webapps/uPortal/': org.objectweb.jonas_web.deployment.api.WebContainerDeploymentDescException:jonas-resource missing for res-ref-name jdbc/PortalDb: jonas-resource missing for res-ref-name jdbc/PortalDb
02:44.29*** join/#tomcat trilluser (~trillian@november.kirovcity.ru)
02:46.22trillusercould anyone answer my question?
02:52.12matroskawhats your question
02:54.42jasonbmatroska: Wow.  That's a whole lot of neat stuff.
02:55.04matroskawhata neat ?
02:56.09trilluserthe thing is:
02:56.09trilluserI use Tomcat 5.
02:56.09trilluserI have the following configuration in server.xml:
02:56.09trilluser...
02:56.09trilluser<Host name="<hostname>" debug="0" appBase="Z:/Clients/Beta/Glendoveertennis" unpackWARs="true" autoDeploy="true">
02:56.11trilluser<Context path="" docBase="Webroot" debug="0" reloadable="true"/>
02:56.13trilluser</Host>
02:56.15trilluserand when tomcat service is started there's error in log:
02:56.17trilluserDocument base Z:\Clients\Beta\Glendoveertennis\Webroot does not exist or is not a readable directory
02:56.19jasonbmatroska: That error message.. it shows how many cool pckages you're running..
02:56.26trilluserZ is a network drive
02:56.56trilluserwhen I change Z: to C: all is fine
02:57.10trilluserand I cannot understand what is a problem?
02:57.21jasonbtrilluser: Sounds like a windows problem.
02:57.41matroskajasonb : so... whats the problem?
02:57.41jasonbtrilluser: What kind of network drive is Z?
02:57.55jasonbtrilluser: Is it an SMB windows file share?
02:58.43jasonbI actually don't know if that should work or not.
02:59.26matroskajasonb: have you deployed uportal or chef from scratch/sources?
03:00.04jasonbmatroska: I haven't even heard of it.  Do you have a URL?
03:00.18trilluserno
03:02.15jasonbHmm.. http://mis105.mis.udel.edu/ja-sig/uportal/
03:02.17trilluserI wonder why does it work on local drive but doesn't work on network drive.
03:02.17trilluserThe network path exists and accessible from Windows
03:02.32matroskachefproject.org and uportal.org
03:02.35jasonbtrilluser: Is it writeable from windows?
03:02.39matroskachef has such a crappy documentation
03:02.46matroskaand the same with uportal
03:03.01trilluserjasonb: yes
03:03.03matroskauportal has better but i still cant get it to run under this jonas
03:03.17matroskamaybe its jonas's problem ... i am running the 4.0 beta
03:03.25matroskath eone that came out two nights ago
03:03.41matroskabut its unlikely that is jonas's problem
03:04.43jasonbmatroska: Ugh, chef builds on Velocity and Turbine.  Don't trust it.
03:05.32jasonbmatroska: Hmm, seems like a deployment/JNDI problem running uportal on Jonas.
03:05.54jasonbmatroska: I've heard lots of good things about Jonas, so I'd be more suspicious of uportal.
03:06.04matroskawhy not?
03:06.24jasonbmatroska: uportal's license looks great.
03:06.37jasonbmatroska: And, looks like lots and lots of people are using it -- another good sign.
03:06.59matroskayes ... but i cant get the stupid thing to run :(
03:07.17matroskafirst its hsql database wont start
03:07.23matroskagives an exception
03:07.48matroskaif i try the mysql version .. it doesnt even want to get deployed - the error above
03:08.11jasonbmatroska: Hmm, looks like uportal doesn't need a whole J2EE server.. it looks (from the docs) as though you can just run it on Tomcat stand-alone.
03:08.21matroskajasonb: .. let me knwo if you can run chef .. i like chef a lot
03:08.35matroskajasonb: i was doing the source build
03:08.40jasonbmatroska: I wish I had time to try all that stuff.  Ugh.
03:08.53matroskai dont like its tomcat standalone
03:09.16matroskait has hsqldb inbuilt ... which doesnt want to run
03:09.27matroskaactually, one has to manually start the stupid hsqldb for uportal
03:09.35matroskaadn the one thats with it doesnt run
03:09.51*** join/#tomcat chsh (~chsh@d235-181-206.home1.cgocable.net)
03:09.51matroskai have like no experience whatsoever with hsql
03:10.41jasonbmatroska: hsql is very nice.  I've used it before, and have only good things to say about it.
03:11.05jasonbmatroska: It's small, pure-Java (so it's portable), and seemingly fast.
03:11.35jasonbmatroska: It's possible for people to bundle hsql and bundle it in a broken way..  That might be the case for uportal, but it's not hsql's fault.
03:12.20matroska<PROTECTED>
03:12.34matroskaeven though it has the .bats they are no goo
03:12.36matroskad
03:12.44jasonbugh.. are you running Windows?
03:13.13matroskayup
03:13.31matroskadont have enough space for linux :(
03:13.33matroskathe entire 40 gigas haddisk is full
03:13.37matroskagigs
03:14.33matroskacan you figure out and tell me how to deploy the tomcat uportal to deploy inside jonas
03:15.28jasonbNo.  Sorry.  I just don't have the time to do stuff like that.
03:15.34matroskaoh
03:16.32jasonbThere are lots of cool apps I'd like to test out on several different application servers, I just don't have the time.
03:17.30jasonbmatroska: But, you should mail one of their mailing lists and tell them the troubles you've had.
03:17.33jasonbhttp://www.uportal.org/getinv.html
03:17.46jasonbMaybe even file a bug.  :)
03:18.40jasonbWelp, gotta go.
03:22.58matroskawait
03:23.02matroskaone last question
03:23.11matroskawhats the default login/password for tomcat5
03:29.12matroskawhats the default login/password for tomcat5