IRC log for #tomcat on 20080508

00:02.14*** join/#tomcat clajo04 (n=clajo04@cpe-72-229-214-63.nyc.res.rr.com)
01:01.56*** join/#tomcat Dwayne_Hicks (n=elsie@adsl-68-74-73-137.dsl.emhril.ameritech.net)
01:16.17*** join/#tomcat odin_ (n=dlm@host86-154-78-206.range86-154.btcentralplus.com) [NETSPLIT VICTIM]
01:16.17*** join/#tomcat andyt (n=andyt@THOMPSON-LAPTOP.ADMIN.SCPS.NYU.EDU) [NETSPLIT VICTIM]
01:16.17*** join/#tomcat mitcivs (n=sv@67.15.117.214) [NETSPLIT VICTIM]
01:16.45*** join/#tomcat jablko (n=jablko@S01060080c6f16ed5.vs.shawcable.net)
01:20.35*** join/#tomcat RurickGrid (n=Erick@200.122.242.159)
01:33.42*** join/#tomcat clajo04_ (n=clajo04_@cpe-72-229-214-63.nyc.res.rr.com)
01:38.24RurickGridhello, anyone know how I can remotely monitoring the jvm
01:39.32*** part/#tomcat clajo04_ (n=clajo04_@cpe-72-229-214-63.nyc.res.rr.com)
01:39.35RurickGridI need to understand the memory leaks
02:29.36*** join/#tomcat INternat (n=nf@123-243-184-161.static.tpgi.com.au)
02:50.32*** join/#tomcat Dwayne_Hicks (n=elsie@adsl-68-74-73-137.dsl.emhril.ameritech.net)
03:13.22*** join/#tomcat acidjnk (n=have@ds87-230-56-32.dedicated.hosteurope.de)
03:24.50*** join/#tomcat markl_ (n=mark@c-24-10-191-137.hsd1.ut.comcast.net)
03:26.50*** join/#tomcat kanai (n=kanai@124.32.114.226)
03:52.46*** join/#tomcat clajo04_ (n=clajo04_@cpe-72-229-214-63.nyc.res.rr.com)
03:54.27*** part/#tomcat clajo04_ (n=clajo04_@cpe-72-229-214-63.nyc.res.rr.com)
03:55.16*** join/#tomcat clajo04_ (n=clajo04_@cpe-72-229-214-63.nyc.res.rr.com)
04:03.46*** join/#tomcat clajo04 (n=clajo04@cpe-72-229-214-63.nyc.res.rr.com)
04:06.08*** join/#tomcat andyt (n=andyt@THOMPSON-LAPTOP.ADMIN.SCPS.NYU.EDU)
05:20.31*** join/#tomcat vikumar (n=Vimal@nat/redhat-in/x-9dfcbbf96eb3b6b1)
06:27.03*** join/#tomcat deebo (n=deebo@stonebay32.com)
06:35.59*** join/#tomcat kjkoster5489 (n=kjkoster@ip188-184-173-82.adsl2.versatel.nl)
06:53.13*** join/#tomcat Dwayne_Hicks (n=elsie@adsl-68-74-73-137.dsl.emhril.ameritech.net)
07:45.59*** join/#tomcat Bollinger (n=peter@88-105-171-42.dynamic.dsl.as9105.com)
08:45.37*** join/#tomcat ponyboy3399 (n=ponyboy3@cpe-72-185-230-234.tampabay.res.rr.com)
08:45.44ponyboy3399anyone there
08:49.50*** part/#tomcat ponyboy3399 (n=ponyboy3@cpe-72-185-230-234.tampabay.res.rr.com)
08:53.40*** join/#tomcat Vanuatoo (n=Vanuatoo@host-85-114-227-35.adsl.caucasus.net)
08:54.45*** join/#tomcat elean (n=elean@90-156-1-90.as.kn.pl)
09:08.10*** join/#tomcat mauriez (i=maui@xs2.xs4all.nl)
10:07.33*** join/#tomcat vikumar (n=Vimal@nat/redhat-in/x-936139d92925e995)
10:34.51*** join/#tomcat lance107 (n=chatzill@061093246097.ctinets.com)
10:35.21lance107i m using 6.0.14......is there any easy way to make catalina logs rotate daily?....thx
10:35.42lance107on linux
10:36.46lance107and tomcat service should not be stopped for log rotation.....i read some docs online for logrotate....but that doesn't fit
10:38.50lance107i m using 6.0.14 on linux.......wanna find some way to do log rotation (not logrotate).....is there any easy way?
10:38.51lance107thx
10:59.38*** join/#tomcat clovisw (n=clovis@201-40-80-179.cscgo300.ipd.brasiltelecom.net.br)
11:00.37kjkoster5489lance107: they already do.
11:00.44kjkoster5489Well, they do on my box.
11:01.07kjkoster5489Are you using a repackaged Tomcat or a vanilla download from apache.org?
11:03.01lance107em.....i got that from apache.org
11:03.25lance107what i got is a big log file.....
11:13.03cloviswhi, i have a problem with Sun JVM, with new Linux Scheduler in a SMP kernel, that make individual threads actually move from processor to processor even though they are in a continuous running state and never sleep, causing cache line migration between the CPUs, this slow down perfomance of tomcat, that in a Single CPU is faster than Dual CPU, any idea?
11:15.25eleanI have deja vu
11:15.35eleanhave you already asked this question ? ;)
11:16.57cloviswyes
11:17.14cloviswbut no good answer, then i try again, ;)
11:17.25cloviswi cant find any solution for that
11:18.01cloviswthe only one, i found, the guy put the tomcat running at a Single CPU machine, its a crazy solution
11:30.01kjkoster5489lance107: check conf/logging.properties, I think you can set these things there.
11:30.51kjkoster5489clovisw: Maybe you can set processor affinity on the process?
11:31.10kjkoster5489The tomcat process, I mean.
11:31.16kjkoster5489Or switch to FreeBSD. :)
11:31.35lance107i know there is a file called logging.properties......but where can i find the related info for that?...e.g. parm for rotation....etc......
11:31.35kjkoster5489Or make Linux not do that. I mean, it's not like this problem is specific to Tomcat.
11:32.07cloviswkjkoster5489: the processor affinity its same solution as to install in a single processor machine
11:32.10kjkoster5489lance107: I think that that file is a bastard log4j config file, so you should be able to use log4j config synta in tin.
11:32.31kjkoster5489clovisw: All processes that have lots of live threads will have this issue.
11:32.46kjkoster5489It sounds like you want to solve a linux kernel issue in Java.
11:33.27kjkoster5489Is there no "#define LEAVE_RUNNING_THREADS_WHERE_THEY_ARE" anywhere?
11:33.30lance107but if i use log4j......i guess it's not java.util.logging......and i need to get another lib from apache......just wondering if tomcat has anything like this come as default pkg....hehe.....and for log4j...em...seems it has some problem for daily rotation....right?
11:34.03kjkoster5489Tomcat uses its own logging package, which was renamed from (I think) log4j.
11:34.11lance107but if u would recommend log4j over tomcat itself....then i will go for it.......just wondering any site has info for log4j properties files parm and detail doc?
11:34.13cloviswkjkoster5489: is this that i'm searching for, ;)
11:34.46kjkoster5489clovisw: downgrade kernel version until this 'innovation' has been exposed for the hype it was and disabled?
11:35.23kjkoster5489Maybe you should ask on a Linuxy channel?
11:35.41kjkoster5489lance107: Google. :)
11:36.39lance107i surely tried that.....kekeke.......but just part of the log4j properties file.....and no walkthru for that.....T.T
11:36.45clovisw<PROTECTED>
11:37.11cloviswbut appears that anyone never get a problem like this
11:42.54kjkoster5489clovisw: maybe Java is the only platform to do serious threading? :)
11:43.20kjkoster5489Have you considered that maybe you need more cpu's, not less to solve this?
11:44.09kjkoster5489That way, Linux will not feel the need to move threads around.
11:44.41clovisw<PROTECTED>
11:45.47kjkoster5489So, as I understand it correctly, the application has to be changed because the OS poops over the CPU caches?
11:45.54kjkoster5489Hmm.
11:46.17kjkoster5489You'd have to run with a hacked Tomcat, a hacked JVM or a hacked SMP kernel.
11:46.42kjkoster5489Maybe the 8-core machine is worth a try. :)
11:48.51cloviswthe slow down performance is progressive, iow, with more core, more slow
11:49.39cloviswi dont believe that this problem, never apperars, no one uses tomcat on linux with SMP?
11:49.55clovisw*apears
11:51.58kjkoster5489I use it on FreeBSD SMP and UP. Never tested UP against SMP for progressive slowdowns, to be honest.
11:52.20kjkoster5489You tested 4-core and 8-core servers?
11:53.15cloviswyes, i used with 4 core, too, same problem
11:53.40kjkoster5489Crap. Otherwise you'd finally have an argument to order that 64-core Sun.
11:54.37kjkoster5489Just to be sure: they are actual cores, not hyperthreaded cpu's?
11:56.30cloviswkjkoster5489: no, its physical cores, i'm sure
11:56.38kjkoster5489Cool, just checking.
12:03.11kjkoster5489How much slowdown do you experience?
12:09.31cloviswabout 40% slow
12:22.15*** join/#tomcat pennsong (n=pennsong@58.247.146.186)
12:27.51pennsonghello everbody, i'm now to tomcat, and I want to install it on my MAC OS X, is it's possible?
12:30.52*** join/#tomcat Woody_ (n=marc@f051167063.adsl.alicedsl.de)
12:35.51Woody_tomcat: 5.5.26-1 java: 6-06-1 os: debian lenny/sid apache 2.2.8-3
12:35.59Woody_I am having this problem: http://rafb.net/p/pM5gKx34.txt
12:37.37Woody_Somehow I don't have the class loaders installed, but I don't know which package they are shipped in, also I don't know if I have the right environmental variable set.
12:39.23Woody_Any help would be great.
12:42.39Woody_My bootstrap.jar is $CATALINA_HOME/bin/bootstrap.jar , but I don't know if the file belongs there or not.
12:47.39*** part/#tomcat pennsong (n=pennsong@58.247.146.186)
12:49.41Woody_I am missing: $CATALINA_HOME/bin/commons-daemon.jar  but have libcommons-daemon-java installed.
13:46.02*** join/#tomcat RurickGrid (n=Erick@201.221.142.243)
14:16.09*** join/#tomcat Infinito_ (n=argos@200-140-155-228.gnace701.dsl.brasiltelecom.net.br)
14:45.47*** join/#tomcat clajo05 (n=clajo04@cpe-72-229-214-63.nyc.res.rr.com)
14:48.56RurickGridI am trying to understand the memory leaks of Tomcat, can help me anybody?
14:49.14*** join/#tomcat clajo06 (n=clajo04@cpe-72-229-214-63.nyc.res.rr.com)
14:51.35*** join/#tomcat Hellaenergy (n=Hellaene@66.187.203.82)
14:53.04*** part/#tomcat Hellaenergy (n=Hellaene@66.187.203.82)
14:59.31*** join/#tomcat eistimax64 (n=user@d01m-195-36-151-235.d4.club-internet.fr)
15:02.16eistimax64does anyone use the build.xml file for Ant from tomcat.apache website?
15:02.28*** part/#tomcat clovisw (n=clovis@201-40-80-179.cscgo300.ipd.brasiltelecom.net.br)
15:03.25*** join/#tomcat Quiddity (n=steve@ng1.cptxoffice.net)
15:06.38*** join/#tomcat riz_ (i=fabrizio@89-96-231-234.ip14.fastwebnet.it)
15:09.09*** join/#tomcat clajo04 (n=clajo04@cpe-72-229-214-63.nyc.res.rr.com)
15:09.31riz_Tomcat 5.5.9 / jrockit-jdk1.5.0_02 /Red Hat Enterprise Linux ES release 3 My question: when I reach the url https://10.30.70.62:8443/wallet/rpc/mfp/getwallettraffic.jspa it ask me for login and password. I would like to change it. How to do it? I see in conf/tomcat-users.xml but I can't see nothing reffered to myuser.
15:10.27riz_Obviously If i try to login three times, it tells me: description This request requires HTTP authentication ().
15:11.45eistimax64Tomcat 6.0.2 / java 1.6 / debian etch - does anyone use the build.xml file for Ant from tomcat.apache website?
15:12.28eistimax64i have a question about the reload target
15:13.50eistimax64the target well update all files in the build directory in my work directory but not in the running tomcat (tomcat/webapps)
15:13.56eistimax64Is it normal?
15:15.07*** part/#tomcat atpa8a (n=stfu@static-71-249-199-57.nycmny.east.verizon.net)
15:15.10*** join/#tomcat randrew (n=raj@dyn-128-59-53-165.dyn.columbia.edu)
15:31.09*** join/#tomcat Vanuatoo (n=Vanuatoo@host-85-114-227-35.adsl.caucasus.net)
16:07.07*** join/#tomcat clajo04 (n=clajo04@cpe-72-229-214-63.nyc.res.rr.com)
16:16.59*** join/#tomcat jasonb (n=jasonb@75.144.23.117)
16:25.01*** part/#tomcat Heikki123 (i=heikki@nappula.net)
16:36.44*** join/#tomcat jablko (n=jablko@216.113.200.202)
17:07.35*** join/#tomcat prgrmr (n=prgrmr@bzq-79-183-139-162.red.bezeqint.net)
17:08.34*** join/#tomcat INternat (n=nf@123-243-184-161.static.tpgi.com.au)
17:27.52*** join/#tomcat eistimax` (n=user@d04m-213-44-60-26.d4.club-internet.fr)
17:37.12*** join/#tomcat Chest (n=elam@d149-67-240-239.try.wideopenwest.com)
17:37.56Chestguys, does the \ escape a quote in the tomcat configuration files?
17:40.46*** join/#tomcat clovisw (n=clovis@201-40-80-179.cscgo300.ipd.brasiltelecom.net.br)
17:41.01jasonbChest: I haven't tested it.. but it might be &quot;
17:46.19cloviswhi, anyone here manage to run cpu affinity per thread basis, on a Linux kernel 2.6 SMP? or managed tomcat run fast on a machine with dual processors than a single processor?
17:51.40*** join/#tomcat yassine (n=yassine@unaffiliated/yassine)
18:00.14jasonbclovisw: There's fast and then there's higher concurrency.  Which one are you after?
18:01.27cloviswjasonb: higher concurrency, many hits
18:02.17jasonbThen you want the OS to schedule Tomcat's threads on as many CPU cores as you have.
18:02.28cloviswjasonb: i have a problem with Sun JVM 1.5.0_15, with the Linux Scheduler in a SMP kernel 2.6, that make individual threads actually move from processor to processor even though they are in a continuous running state and never sleep, causing cache line migration between the CPUs, this slow down perfomance of multithread java applications, that in a Single CPU is faster than Dual CPU
18:03.36jasonbclovisw: How are you doing performance testing such that you know it is slower on a single CPU machine?
18:03.59cloviswjasonb: its by processor basis
18:04.25jasonbclovisw: If you want better integration with the kernel, and better overall performance, you must stop using Java 1.5 and move to 1.6 or 1.7.
18:04.56cloviswjasonb: a single processor Xeon, can process for example: 400 request, on a dual processor, same power, but dual, process is less requests
18:06.20cloviswjasonb: are you sure that jvm 1.6 manage the new linux Scheduler better?
18:07.18cloviswjasonb: with the jvm 1.6 in some serialization, i'm getting stack overflow, that i dont get with 1.5, specially complex graph objects
18:07.23jasonbclovisw: Newer JVMs offer all kinds of improvements over the older ones, including better integration with the OS and better performance.
18:08.01jasonbclovisw: You've tried only the Sun Java 1.6?
18:08.15jasonbclovisw: Also try the BEA JRockit 1.6.
18:08.28cloviswjasonb: then i must make a downgrade to last version of 1.5, but i get probem with threads
18:08.56cloviswjasonb: i tried to find BEA downloads, but dont have the download anymore
18:11.47*** part/#tomcat Chest (n=elam@d149-67-240-239.try.wideopenwest.com)
18:13.04jasonbYeah, I guess they're transitioning their web pages to Oracle's control right now.
18:13.36cloviswjasonb: then for now JRockit isn't a option, i will try the IBM ones then
18:13.58jasonbYeah, I think IBM has a 1.6 JDK.
18:16.00*** join/#tomcat dbaio (n=danilo@unaffiliated/dbaio)
18:17.27dbaiohey guys... in my tomcat manager i have many virtual hosts... and theses virtual host have a /_manager   ... but  the manager list all applications of the tomcat... and all the tomcat home folders.....    can i show onlye the virtualhost application ?
18:17.44cloviswjasonb: i just downloading
18:19.13*** join/#tomcat elean (n=elean@90-156-1-90.as.kn.pl)
18:23.41cloviswjasonb: i hope that serialization between Sun JVM and IBM JVM, works, ;)
18:25.26jasonbclovisw: It's JVM implementation specific, so it probably doesn't.
18:26.06*** join/#tomcat Dwayne_Hicks (n=elsie@adsl-68-74-73-137.dsl.emhril.ameritech.net)
18:28.04cloviswjasonb: then i have no option, i must use Sun JVM
18:28.47cloviswjasonb: maybe i migrate to Solaris? appears that in Solaris dont have this type of problem with Sun JVM
18:30.07jasonbI can't really suggest doing that.. It could work, but then you'd likely have worse problems.
18:36.48*** join/#tomcat eistimax` (n=user@d04m-212-194-200-8.d4.club-internet.fr)
18:41.37*** join/#tomcat billd (n=SA-ENG@wsip-68-105-243-3.sd.sd.cox.net)
18:57.29*** join/#tomcat npm (n=NielsMay@cpe-75-83-171-237.socal.res.rr.com)
19:51.49Woody_tomcat: 5.5.26-1 java: 6-06-1 os: debian lenny/sid apache 2.2.8-3
19:52.38Woody_I am having this problem: http://rafb.net/p/pM5gKx34.txt , but I disabled the security manager, I have know Idea why it still tells me I didn't.
19:53.14Woody_I am missing: $CATALINA_HOME/bin/commons-daemon.jar  but have libcommons-daemon-java installed.
19:53.55Woody_Any help would be great.
19:55.09*** join/#tomcat gregor_k (n=a@stgt-4dbdb152.pool.einsundeins.de)
19:56.02*** join/#tomcat eistimax` (n=user@d02m-213-44-187-120.d4.club-internet.fr)
20:32.09*** join/#tomcat momelod (n=smelo@CPE00a065c98ce6-CM0012c91df0bc.cpe.net.cable.rogers.com)
20:32.13momelodgreetings channel
20:32.28eleangreetings newcomer
20:32.31momelodfor some reason, i cannot undeploy an app with tomcat manager
20:32.37momelod:) hi elean
20:33.04eleanwhat do you mean by "can't" ?
20:33.14eleanjar's are not being deleted ?
20:33.17eleanjars
20:33.41momelodi mean, it stops the app, and removes most of the files.. it leaves files *.lib
20:33.53momelodbah i mean, lib/*.jar
20:34.07eleanit happens
20:34.14eleanactually I have this problem myself
20:34.19momelodshoot
20:34.24eleanand have no idea how to solve it neatly
20:34.36momeloddo u have a dirty hack?
20:34.36eleanone solution is to set antiResourceLocking = true
20:34.42eleanin Context element
20:34.51eleanbut it has the nasty shortcoming ..
20:34.56momelodwhat's that?
20:35.07eleanjsp's are not reloaded so you can't really use this setting in dev mode
20:35.14eleangood for production tough
20:37.41eleanhmm
20:38.00eleanI've just noticed there is an 'antiJARLocking' attribute as well ..
20:38.19eleanpossible this one would work
20:39.45*** join/#tomcat cofeineSunshine (n=justinas@tavatievs.eu)
20:42.31momelodany side effects with antiJARLocking?
20:43.07eleanhttp://tomcat.apache.org/tomcat-5.5-doc/config/context.html
20:43.11eleannothing is specified there
20:43.21eleanbut I don't even know if it works
20:43.31momelodhttp://forum.springframework.org/archive/index.php/t-24480.html
20:43.39momelodsome1 here describes the same solution
20:44.20eleanyeap but he advised to include both attrs
20:44.33eleanI'm curious if antiJARlocking alone would work ..
20:44.37eleanit'd solve my problems also ;)
20:44.43eleangotta give it try later
20:44.47eleana try
20:45.29elean"This problems appears on Windows" - another reason why I should immediately move to linux :)
20:46.10jasonbLinux works quite well for Tomcat..  for me anyway.
20:47.58cloviswjasonb: a guy from Suns Forum, reply me with "Sun's JVM (HotSpot) doesn't do any thread scheduling, it leaves it up to the operating system. So the o/s is choosing to migrate threads. The obvious things for slowing down a multi-threaded app are lock contention or possibly garbage collection (multiple CPUs allow more work to get done, which means more heap allocation and more frequent GCs if using the same heap size as a sin
20:48.37cloviswjasonb: since there isnt a GC thing, maybe a lock, in logs, like accessValve?
20:48.53jasonbclovisw: Yeah.  Sounds right.  It's pretty hard to compare single processing performance to multiprocessing performance.
20:49.13jasonbclovisw: It could very well be a GC thing.
20:49.25cloviswjasonb: since all http thread must write to access log file
20:50.13jasonbTry turning off the access log valve temporarily and see how that changes the performance (I don't expect it to suddenly become identical to the performance of a single processor machine though).
20:50.38jasonbclovisw: You're really comparing apples and oranges.
20:52.58momelodelean: i tried with just antiJARLocking
20:53.03momelodstill didn't work
20:53.26eleanmomelod: pity ..
20:53.49eleanmomelod: I'd rather have me deployment suck than loose jsp reloading..
21:05.44*** part/#tomcat clovisw (n=clovis@201-40-80-179.cscgo300.ipd.brasiltelecom.net.br)
21:05.52*** part/#tomcat Tuju (n=tuju@50.157.190.90.sta.estpak.ee)
21:14.57*** join/#tomcat a4akba (n=a4akb@62.215.156.215)
21:38.16*** join/#tomcat balla (n=balla@adsl-ull-199-43.51-151.net24.it)
21:44.04*** join/#tomcat que (n=a4akb@62.215.156.215)
22:01.22*** join/#tomcat Infinito_ (n=argos@200-140-155-228.gnace701.dsl.brasiltelecom.net.br)
23:12.03*** join/#tomcat bill1 (n=SA-ENG@wsip-68-105-243-3.sd.sd.cox.net)
23:12.15*** join/#tomcat Dwayne_Hicks (n=elsie@adsl-68-74-73-137.dsl.emhril.ameritech.net)

Generated by irclog2html.pl Modified by Tim Riker to work with infobot.