July 2006 - jboss-dev-forums - Jboss List Archives

[Design of JBoss/Tomcat Integration] - Tomcat performance bottleneck

by michael.yuan＠jboss.com

Hi guys, This is not really a jboss/tomcat integration issue. But I cannot find a better forum to discuss it ... So, here it goes: In my performance tests at Dell, I identified a likely performance contention point in Tomcat. The symptom is that we cannot drive the server CPU utilization beyond 90 percent. A quick thread dump shows that most threads are waiting to execute a synchronized HashMap access block in line 175 of the org.apache.catalina.core.ApplicationContext class: | public class ApplicationContext | implements ServletContext { | | ... ... | | public Object getAttribute(String name) { | | synchronized (attributes) { | return (attributes.get(name)); | } | | } | } | Since the server has multiple CPUs, they must wait in turn to run this block of code and hence results in lower CPU utilization. The best solution is to use a concurrent hashmap to get rid of the synchronized block. Can we get this fixed? Thanks. I attach a more detailed stack trace from the thread dump below. | "http-0.0.0.0-8080-17" daemon prio=1 tid=0x0000002c918799d0 | nid=0x107b waiting for monitor entry | [0x0000000044c9a000..0x0000000044c9bc90] | at org.apache.catalina.core.ApplicationContext.getAttribute(ApplicationContext.java:175) | - waiting to lock <0x0000002ae2a4f980> (a java.util.HashMap) | at org.apache.catalina.core.ApplicationContextFacade.getAttribute(ApplicationContextFacade.java:315) | at org.apache.myfaces.context.servlet.ApplicationMap.getAttribute(ApplicationMap.java:41) | at org.apache.myfaces.context.servlet.AbstractAttributeMap.get(AbstractAttributeMap.java:87) | at org.jboss.seam.contexts.FacesApplicationContext.get(FacesApplicationContext.java:46) | at org.jboss.seam.Component.forName(Component.java:1138) | at org.jboss.seam.ejb.SeamInterceptor.getSeamComponent(SeamInterceptor.java:93) | at org.jboss.seam.ejb.SeamInterceptor.aroundInvokeInContexts(SeamInterceptor.java:66) | at org.jboss.seam.ejb.SeamInterceptor.aroundInvoke(SeamInterceptor.java:45) | at sun.reflect.GeneratedMethodAccessor85.invoke(Unknown Source) | at sun.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:25) | at java.lang.reflect.Method.invoke(Method.java:585) | at org.jboss.ejb3.interceptor.InvocationContextImpl.proceed(InvocationContextImpl.java:118) | at org.jboss.ejb3.interceptor.EJB3InterceptorsInterceptor.invoke(EJB3InterceptorsInterceptor.java:63) | at org.jboss.aop.joinpoint.MethodInvocation.invokeNext(MethodInvocation.java:101) | at org.jboss.ejb3.entity.ExtendedPersistenceContextPropagationInterceptor.invoke(ExtendedPersistenceContextPropagationInterceptor.java:57) | at org.jboss.aop.joinpoint.MethodInvocation.invokeNext(MethodInvocation.java:101) | at org.jboss.ejb3.entity.TransactionScopedEntityManagerInterceptor.invoke(TransactionScopedEntityManagerInterceptor.java:54) | at org.jboss.aop.joinpoint.MethodInvocation.invokeNext(MethodInvocation.java:101) | at org.jboss.ejb3.AllowedOperationsInterceptor.invoke(AllowedOperationsInterceptor.java:47) | at org.jboss.aop.joinpoint.MethodInvocation.invokeNext(MethodInvocation.java:101) | at org.jboss.aspects.tx.TxPolicy.invokeInOurTx(TxPolicy.java:79) | at org.jboss.aspects.tx.TxInterceptor$Required.invoke(TxInterceptor.java:197) | at org.jboss.aop.joinpoint.MethodInvocation.invokeNext(MethodInvocation.java:101) | at org.jboss.aspects.tx.TxPropagationInterceptor.invoke(TxPropagationInterceptor.java:76) | at org.jboss.aop.joinpoint.MethodInvocation.invokeNext(MethodInvocation.java:101) | at org.jboss.ejb3.stateful.StatefulInstanceInterceptor.invoke(StatefulInstanceInterceptor.java:81) | at org.jboss.aop.joinpoint.MethodInvocation.invokeNext(MethodInvocation.java:101) | at org.jboss.aspects.security.AuthenticationInterceptor.invoke(AuthenticationInterceptor.java:77) | at org.jboss.aop.joinpoint.MethodInvocation.invokeNext(MethodInvocation.java:101) | at org.jboss.ejb3.ENCPropagationInterceptor.invoke(ENCPropagationInterceptor.java:47) | at org.jboss.aop.joinpoint.MethodInvocation.invokeNext(MethodInvocation.java:101) | at org.jboss.ejb3.asynchronous.AsynchronousInterceptor.invoke(AsynchronousInterceptor.java:106) | at org.jboss.aop.joinpoint.MethodInvocation.invokeNext(MethodInvocation.java:101) | at org.jboss.ejb3.stateful.StatefulContainer.localInvoke(StatefulContainer.java:180) | at org.jboss.ejb3.stateful.StatefulLocalProxy.invoke(StatefulLocalProxy.java:98) | at $Proxy72.getCart(Unknown Source) | View the original post : http://www.jboss.com/index.html?module=bb&op=viewtopic&p=3960546#3960546 Reply to the post : http://www.jboss.com/index.html?module=bb&op=posting&mode=reply&p=3960546

19 years, 8 months

1
0
0 / 0

[Design of JBoss jBPM] - Re: Cancelling task signals token over default transition.

by kukeltje

Chris, imo it is an oversight. I can't think of a good reason. What would you think of having an .cancel() take the default transition, having and .cancel(Transition trans) or .cancel(String transName) take the one specified. The required .end(...) methods are already available. Not taking any transition is a difficult case. If this is the only task, the process would in in undefined state. If this task(node) is one of the teeth(?) of a fork, the join needs the token, so it has to take a transition for the join to work. Only if there are multiple tasks in a node this is an option. Let's leave this one out for the moment View the original post : http://www.jboss.com/index.html?module=bb&op=viewtopic&p=3960528#3960528 Reply to the post : http://www.jboss.com/index.html?module=bb&op=posting&mode=reply&p=3960528

19 years, 8 months

1
0
0 / 0

[Design of JBoss jBPM] - Re: JPDL Eclipse Graphical Editor - problem building.

by koen.aers＠jboss.com

You don't need 'a' zip tool on the command line or 'a' cvs client, but the exact same zip tool and cvs client as specified by the Eclipse PDE build guidelines. So I think we should collectively booh the Eclipse PDE team on this one indeed... Nevertheless, I have some good news that Tom will like a lot. I have worked my way around the requirement to checkout the plugin and feature projects during the Eclipse build. If you checkout the entire jbpm.3 tree and invoke the build.xml in the build project, no cvs checkout will happen during the GPD build. So it becomes faster and faster. What currently takes most of the time is the Eclipse installation. I will see what I can do to work around this too so that it only happens when the Eclipse configuration has changed. Regards, Koen View the original post : http://www.jboss.com/index.html?module=bb&op=viewtopic&p=3960482#3960482 Reply to the post : http://www.jboss.com/index.html?module=bb&op=posting&mode=reply&p=3960482

19 years, 8 months

1
0
0 / 0

[Design of JBoss Collaboration Server] - Re: JBMS Alternate Datastore

by csand101

acoliver, Not sure what you are referring to by a non-free HA implementation. All I see at MySQL.com is info on the MySQL cluster architectur (like the white paper "A Guide to High Availability Clustering: How MySQL Supports 99.999% Availability" and of course the clustering section of the docs). Each node entry point (MySQL server) has it's own ip. No mention is made of MySQL servers taking over ips which would require a single MySQL server (the takeover server) to listen to multiple ips at once (not to mention that now one server has to handle twice it's load, even if 5 nodes in cluster). There is mention however of "applications" failing over to working MySQL nodes which seems like the most sensible solution. 4 nodes split the load of the downed 5th node. Tell me, how can JBMS be such a great enterprise mail server if it can't failover to another datastore instance, which it relies on to store mail? There is a single point of failure at the JDBC connection layer, at least for that JBMS node. And what happens when a JBMS node can't reach it's database? Does it stop listening for POP3 and SMTP connections (which would at least allow a load balancer in front of the mail server to direct traffic to another JBMS server) or does it just keep listening for connections and replying with errors (requiring multiple MX records so the sending mail server can use a different one)? Worse yet, it would keep spooling mail and throwing SQLExceptions becasuse the database is gone. Any light you could shed on this issue, acoliver, would be appreciated... surely by more than just sappenin and I as I don't see any JBMS documentation on this issue. Also, please correct me if I'm wrong anywhere above. Especially if it means there is not actually a free HA JBMS solution. Thank you View the original post : http://www.jboss.com/index.html?module=bb&op=viewtopic&p=3960471#3960471 Reply to the post : http://www.jboss.com/index.html?module=bb&op=posting&mode=reply&p=3960471

19 years, 8 months

1
0
0 / 0

[Design of Security on JBoss] - Re: Authorization framework

by anil.saldhana＠jboss.com

Latest Questions: 1) Shall we retire the old custom Sun parser parsed login-config.xml in the conf directory to a JBossXB based login-config.xml (security_config_5_0.xsd)? 2) Should we update the 5.0 schema such that there is always atleast an authentication element or an authorization element? Currently, we have mandatory authentication element, with an optional authorization element. View the original post : http://www.jboss.com/index.html?module=bb&op=viewtopic&p=3960452#3960452 Reply to the post : http://www.jboss.com/index.html?module=bb&op=posting&mode=reply&p=3960452

19 years, 8 months

1
0
0 / 0

[Design of JBoss Collaboration Server] - Re: JBMS Alternate Datastore

by sappenin

Any idea how something like this would work with Postgresql? View the original post : http://www.jboss.com/index.html?module=bb&op=viewtopic&p=3960441#3960441 Reply to the post : http://www.jboss.com/index.html?module=bb&op=posting&mode=reply&p=3960441

19 years, 8 months

1
0
0 / 0

[Design of JBoss jBPM] - Cancelling task signals token over default transition.

by cpob

See jira issue - http://jira.jboss.com/jira/browse/JBPM-710 and User Forums thread http://www.jboss.com/index.html?module=bb&op=viewtopic&p=3960424#3960424 In short, when you cancel a task, it calls .end(); This calls the default transition. kukeltje recommended I reported it here as it might be seen by more developers. Is there any specific reason why a cancel does this, or is it an oversight? View the original post : http://www.jboss.com/index.html?module=bb&op=viewtopic&p=3960426#3960426 Reply to the post : http://www.jboss.com/index.html?module=bb&op=posting&mode=reply&p=3960426

19 years, 8 months

1
0
0 / 0

[Design of JBoss Collaboration Server] - Re: Agregar un MAilListener

by acoliver＠jboss.org

You must either add some management interface: MyClass implements MyClassMBean OR create an xmbean descriptor. (Look at one of the mbeans with xmbean-dd= in the jboss-service.xml and look at the mentioned file) The error means that you have neither. View the original post : http://www.jboss.com/index.html?module=bb&op=viewtopic&p=3960413#3960413 Reply to the post : http://www.jboss.com/index.html?module=bb&op=posting&mode=reply&p=3960413

19 years, 8 months

1
0
0 / 0

[Design of JBoss Collaboration Server] - Re: JBMS Alternate Datastore

by acoliver＠jboss.org

MySQL has an HA option. It is not free though. A load balancer is not sufficient. Go to MySQL.com for more details. View the original post : http://www.jboss.com/index.html?module=bb&op=viewtopic&p=3960409#3960409 Reply to the post : http://www.jboss.com/index.html?module=bb&op=posting&mode=reply&p=3960409

19 years, 8 months

1
0
0 / 0

[Design of Kosmos] - Re: serviceResultUpdaterJob not traversing repository

by sstuart4

Thanks Aron, that makes much more sense. Maybe this sort of explanation on the kosmos site would be helpful for future users. The information on setting up a job for traversing a repository does not give any indication of how/when the initial caching is supposed to work. View the original post : http://www.jboss.com/index.html?module=bb&op=viewtopic&p=3960398#3960398 Reply to the post : http://www.jboss.com/index.html?module=bb&op=posting&mode=reply&p=3960398

19 years, 8 months

1
0
0 / 0

2026

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

jboss-dev-forums July 2006