November 2007 - jboss-dev-forums - Jboss List Archives

[Design of Clustering on JBoss (Clusters/JBoss)] - Re: JBAS-4960 - broken merge

by bstansberry＠jboss.com

Yeah, no excuse there. I think I'm going to hack my svn client to disallow commits between 6:00 PM Friday and 9:00 AM Monday. The risk/reward ratio is just too high. View the original post : http://www.jboss.com/index.html?module=bb&op=viewtopic&p=4103695#4103695 Reply to the post : http://www.jboss.com/index.html?module=bb&op=posting&mode=reply&p=4103695

18 years, 4 months

1
0
0 / 0

[Design of Clustering on JBoss (Clusters/JBoss)] - Re: JBAS-4960 - broken merge

by adrian＠jboss.org

I think you must have forgetten to commit it to repository.jbosa.com? :-) But the merge problem was causing one of the tests to fail in the metadata project itself. View the original post : http://www.jboss.com/index.html?module=bb&op=viewtopic&p=4103691#4103691 Reply to the post : http://www.jboss.com/index.html?module=bb&op=posting&mode=reply&p=4103691

18 years, 4 months

1
0
0 / 0

[Design the new POJO MicroContainer] - Re: KernelDeployment from -beans.properties

by alesj

"adrian(a)jboss.org" wrote : Have a look at the log.properties format. This is a property file equivalent of xml log4j.properties? For simple javabeans props this is trvial/natural. I was thinking more about class, constructor, lifecycle (create, start, stop, destroy), installs, callbacks, ... Currently I do something like this: mybean.(class)=org.jboss.acme.MyBean mybean.(start)=someStartMethod ... Everything after bean name that is not in brackets () is considered as a property. Which is what you will mostly want. View the original post : http://www.jboss.com/index.html?module=bb&op=viewtopic&p=4103685#4103685 Reply to the post : http://www.jboss.com/index.html?module=bb&op=posting&mode=reply&p=4103685

18 years, 4 months

1
0
0 / 0

[Design of Clustering on JBoss (Clusters/JBoss)] - Re: JBAS-4960 - broken merge

by bstansberry＠jboss.com

Crap. I'm sorry, Adrian; :-( I'd thought I'd tested this against a clean AS checkout, but it seems it wasn't clean. View the original post : http://www.jboss.com/index.html?module=bb&op=viewtopic&p=4103681#4103681 Reply to the post : http://www.jboss.com/index.html?module=bb&op=posting&mode=reply&p=4103681

18 years, 4 months

1
0
0 / 0

[Design the new POJO MicroContainer] - Re: KernelDeployment from -beans.properties

by adrian＠jboss.org

"alesj" wrote : I've created this JIRA issue: | - http://jira.jboss.com/jira/browse/JBMICROCONT-217 | | And I've already commited some initial code - in order to not loose it while I'm doing other things. :-) | The idea that I had is to crawl up and down the tree and along the way create appropriate metadata - see current implementation. | | What kind of semantics should we use to define bean metadata? | Any special request or directions regarding analogy with XML? | | Have a look at the log.properties format. This is a property file equivalent of xml View the original post : http://www.jboss.com/index.html?module=bb&op=viewtopic&p=4103675#4103675 Reply to the post : http://www.jboss.com/index.html?module=bb&op=posting&mode=reply&p=4103675

18 years, 4 months

1
0
0 / 0

[Design of Clustering on JBoss (Clusters/JBoss)] - Re: JBAS-4960 - broken merge

by adrian＠jboss.org

I also had to "fix" the Tomcat deployer. Can you check this patch? | [ejort@warjort tomcat]$ svn diff | Index: src/main/org/jboss/web/tomcat/service/deployers/TomcatDeployer.java | =================================================================== | --- src/main/org/jboss/web/tomcat/service/deployers/TomcatDeployer.java (revision 66950) | +++ src/main/org/jboss/web/tomcat/service/deployers/TomcatDeployer.java (working copy) | @@ -769,11 +769,12 @@ | if (metaData.getDistributable() != null) | { | ReplicationConfig repCfg = metaData.getReplicationConfig(); | - repCfg.setDefaultCacheName(this.cacheName); | - repCfg.setDefaultUseJK(this.useJK); | - repCfg.setDefaultSnapshotMode(this.snapshotMode); | - repCfg.setDefaultSnapshotInterval(this.snapshotInterval); | - repCfg.setDefaultUseLocalCache(this.useLocalCache); | + repCfg.setCacheName(this.cacheName); | + repCfg.setUseJKBoolean(this.useJK); | + repCfg.setSnapshotMode(this.snapshotMode); | + repCfg.setSnapshotIntervalInteger(this.snapshotInterval); | + // This appears to be not used anymore??? - See JBAS-4960 | + //repCfg.setUseLocalCache(this.useLocalCache); | } | | //Check if there are any xacml policy files | View the original post : http://www.jboss.com/index.html?module=bb&op=viewtopic&p=4103672#4103672 Reply to the post : http://www.jboss.com/index.html?module=bb&op=posting&mode=reply&p=4103672

18 years, 4 months

1
0
0 / 0

[Design the new POJO MicroContainer] - KernelDeployment from -beans.properties

by alesj

I've created this JIRA issue: - http://jira.jboss.com/jira/browse/JBMICROCONT-217 And I've already commited some initial code - in order to not loose it while I'm doing other things. :-) The idea that I had is to crawl up and down the tree and along the way create appropriate metadata - see current implementation. What kind of semantics should we use to define bean metadata? Any special request or directions regarding analogy with XML? View the original post : http://www.jboss.com/index.html?module=bb&op=viewtopic&p=4103670#4103670 Reply to the post : http://www.jboss.com/index.html?module=bb&op=posting&mode=reply&p=4103670

18 years, 4 months

1
0
0 / 0

[Design of Clustering on JBoss (Clusters/JBoss)] - Re: Related to JBAS-4950: Exception handling in UnifiedInvok

by bstansberry＠jboss.com

anonymous wrote : My intention here is to determine what's missing or needs changing in Unified proxy so that it behaves the same way as the JRMP when it comes to transaction handling and tx failover authorisations. Please use the opportunity to think a bit about whether there are any holes in JRMP as well. That code's been looked over a lot more, so less likely to have issues, but another quick review can't hurt. anonymous wrote : I will probably create a different jira for this, rather than using JBAS-4950, because this jira focuses on the calling invocationHasReachedAServer() upon successfull returns, rather than exception handling itself. Thoughts? Yes, this is a separate issue. View the original post : http://www.jboss.com/index.html?module=bb&op=viewtopic&p=4103665#4103665 Reply to the post : http://www.jboss.com/index.html?module=bb&op=posting&mode=reply&p=4103665

18 years, 4 months

1
0
0 / 0

[Design of Clustering on JBoss (Clusters/JBoss)] - Related to JBAS-4950: Exception handling in UnifiedInvokerHA

by galder.zamarreno＠jboss.com

Re: http://jira.jboss.com/jira/browse/JBAS-4950 The fix for this issue is pretty simple, but I feel there's not enough testing in this area to make sure that the port from JRMP invoker to Unified invoker proxy was done correctly. I'm working on a series of tests for JBAS-4950 and JBAS-4455 to solve these shortcomings. There's a topic in the JRMP/Unified proxies that requires further thought, which is exception handling. Let's compare the two: 1.- JRMPInvokerProxyHA Exception handling that does not lead to invocationHasReachedAServer() being called: * java.net.ConnectException * java.net.UnknownHostException * java.rmi.ConnectException * java.rmi.ConnectIOException * java.rmi.NoSuchObjectException * java.rmi.UnknownHostException * GenericClusteringException.NO Exception handling that leads to invocationHasReachedAServer() being called: * GenericClusteringException.YES * java.rmi.ServerException * java.lang.Exception 2.- UnifiedInvokerHAProxy Exception handling that does not lead to invocationHasReachedAServer() being called: * org.jboss.remoting.CannotConnectException * GenericClusteringException.NO * java.rmi.RemoteException * java.lang.Throwable Exception handling that leads to invocationHasReachedAServer() being called: * GenericClusteringException.YES As you can see, there're noticeable differences between the handling of the two proxies, and we have no tests to validate that the port from JRMP to Unified was done correctly. My intention here is to determine what's missing or needs changing in Unified proxy so that it behaves the same way as the JRMP when it comes to transaction handling and tx failover authorisations. I will probably create a different jira for this, rather than using JBAS-4950, because this jira focuses on the calling invocationHasReachedAServer() upon successfull returns, rather than exception handling itself. Thoughts? View the original post : http://www.jboss.com/index.html?module=bb&op=viewtopic&p=4103659#4103659 Reply to the post : http://www.jboss.com/index.html?module=bb&op=posting&mode=reply&p=4103659

18 years, 4 months

1
0
0 / 0

[Design of Clustering on JBoss (Clusters/JBoss)] - Re: JBAS-4960 - broken merge

by adrian＠jboss.org

I "fixed it" by changing the code to read: | if(override != null && override.maxActiveSessions != null && override.maxActiveSessions != -1) | setMaxActiveSession(override.maxActiveSessions); | View the original post : http://www.jboss.com/index.html?module=bb&op=viewtopic&p=4103649#4103649 Reply to the post : http://www.jboss.com/index.html?module=bb&op=posting&mode=reply&p=4103649

18 years, 4 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 November 2007