[mod_cluster-issues] [JBoss JIRA] (MODCLUSTER-321) mod_cluster doesn't properly use timeouts in initial proxyList connection attempts

[ https://issues.jboss.org/browse/MODCLUSTER-321?page=com.atlassian.jira.pl... ] Aaron Ogburn updated MODCLUSTER-321: ------------------------------------ Description: mod_cluster doesn't properly use timeouts in initial proxyList connection attempts when JBoss is starting. This can delay start up by several minutes at a time as the main thread stalls in the following call: "main" prio=10 tid=0x000000004318d800 nid=0x4085 runnable [0x0000000041822000] java.lang.Thread.State: RUNNABLE at java.net.PlainSocketImpl.socketConnect(Native Method) at java.net.PlainSocketImpl.doConnect(PlainSocketImpl.java:333) - locked <0x00000000f093e6a8> (a java.net.SocksSocketImpl) at java.net.PlainSocketImpl.connectToAddress(PlainSocketImpl.java:195) at java.net.PlainSocketImpl.connect(PlainSocketImpl.java:182) at java.net.SocksSocketImpl.connect(SocksSocketImpl.java:366) at java.net.Socket.connect(Socket.java:529) at java.net.Socket.connect(Socket.java:478) at java.net.Socket.<init>(Socket.java:375) at java.net.Socket.<init>(Socket.java:218) at javax.net.DefaultSocketFactory.createSocket(SocketFactory.java:212) at org.jboss.modcluster.mcmp.impl.DefaultMCMPHandler$Proxy.getConnection(DefaultMCMPHandler.java:1141) - locked <0x00000000f08aa980> (a org.jboss.modcluster.mcmp.impl.DefaultMCMPHandler$Proxy) at org.jboss.modcluster.mcmp.impl.DefaultMCMPHandler$Proxy.getConnectionWriter(DefaultMCMPHandler.java:1166) - locked <0x00000000f08aa980> (a org.jboss.modcluster.mcmp.impl.DefaultMCMPHandler$Proxy) at org.jboss.modcluster.mcmp.impl.DefaultMCMPHandler.sendRequest(DefaultMCMPHandler.java:776) at org.jboss.modcluster.mcmp.impl.DefaultMCMPHandler.sendRequest(DefaultMCMPHandler.java:875) - locked <0x00000000f08aa980> (a org.jboss.modcluster.mcmp.impl.DefaultMCMPHandler$Proxy) at org.jboss.modcluster.mcmp.impl.DefaultMCMPHandler.status(DefaultMCMPHandler.java:465) at org.jboss.modcluster.mcmp.impl.DefaultMCMPHandler.init(DefaultMCMPHandler.java:140) at org.jboss.modcluster.CatalinaEventHandler.init(CatalinaEventHandler.java:113) - locked <0x00000000ffdbaac0> (a org.jboss.modcluster.CatalinaEventHandler) at org.jboss.modcluster.CatalinaEventHandler.init(CatalinaEventHandler.java:56) at org.jboss.modcluster.CatalinaEventHandlerAdapter.lifecycleEvent(CatalinaEventHandlerAdapter.java:127) at org.jboss.modcluster.AbstractModClusterService.lifecycleEvent(AbstractModClusterService.java:90) at org.jboss.web.tomcat.service.deployers.MicrocontainerIntegrationLifecycleListener.lifecycleEvent(MicrocontainerIntegrationLifecycleListener.java:102) at org.apache.catalina.util.LifecycleSupport.fireLifecycleEvent(LifecycleSupport.java:117) at org.apache.catalina.core.StandardServer.start(StandardServer.java:741) at org.apache.catalina.startup.Catalina.start(Catalina.java:575) This happens because DefaultMCMPHandler$Proxy.getConnection hands the address/port straight to the socket factory and so Socket.connect is called without any timeout. Instead, DefaultSocketFactory.createSocket should be called without a given ip/port so that the returned Socket can then be bound with a proper timeout through Socket.connect(SocketAddress endpoint, int timeout). mod_cluster 1.2 currently already does this. was: mod_cluster doesn't properly use timeouts in initial proxyList connection attempts when JBoss is starting. This can delay start up by several minutes at a time as the main thread stalls in the following call: "main" prio=10 tid=0x000000004318d800 nid=0x4085 runnable [0x0000000041822000] java.lang.Thread.State: RUNNABLE at java.net.PlainSocketImpl.socketConnect(Native Method) at java.net.PlainSocketImpl.doConnect(PlainSocketImpl.java:333) - locked <0x00000000f093e6a8> (a java.net.SocksSocketImpl) at java.net.PlainSocketImpl.connectToAddress(PlainSocketImpl.java:195) at java.net.PlainSocketImpl.connect(PlainSocketImpl.java:182) at java.net.SocksSocketImpl.connect(SocksSocketImpl.java:366) at java.net.Socket.connect(Socket.java:529) at java.net.Socket.connect(Socket.java:478) at java.net.Socket.<init>(Socket.java:375) at java.net.Socket.<init>(Socket.java:218) at javax.net.DefaultSocketFactory.createSocket(SocketFactory.java:212) at org.jboss.modcluster.mcmp.impl.DefaultMCMPHandler$Proxy.getConnection(DefaultMCMPHandler.java:1141) - locked <0x00000000f08aa980> (a org.jboss.modcluster.mcmp.impl.DefaultMCMPHandler$Proxy) at org.jboss.modcluster.mcmp.impl.DefaultMCMPHandler$Proxy.getConnectionWriter(DefaultMCMPHandler.java:1166) - locked <0x00000000f08aa980> (a org.jboss.modcluster.mcmp.impl.DefaultMCMPHandler$Proxy) at org.jboss.modcluster.mcmp.impl.DefaultMCMPHandler.sendRequest(DefaultMCMPHandler.java:776) at org.jboss.modcluster.mcmp.impl.DefaultMCMPHandler.sendRequest(DefaultMCMPHandler.java:875) - locked <0x00000000f08aa980> (a org.jboss.modcluster.mcmp.impl.DefaultMCMPHandler$Proxy) at org.jboss.modcluster.mcmp.impl.DefaultMCMPHandler.status(DefaultMCMPHandler.java:465) at org.jboss.modcluster.mcmp.impl.DefaultMCMPHandler.init(DefaultMCMPHandler.java:140) at org.jboss.modcluster.CatalinaEventHandler.init(CatalinaEventHandler.java:113) - locked <0x00000000ffdbaac0> (a org.jboss.modcluster.CatalinaEventHandler) at org.jboss.modcluster.CatalinaEventHandler.init(CatalinaEventHandler.java:56) at org.jboss.modcluster.CatalinaEventHandlerAdapter.lifecycleEvent(CatalinaEventHandlerAdapter.java:127) at org.jboss.modcluster.AbstractModClusterService.lifecycleEvent(AbstractModClusterService.java:90) at org.jboss.web.tomcat.service.deployers.MicrocontainerIntegrationLifecycleListener.lifecycleEvent(MicrocontainerIntegrationLifecycleListener.java:102) at org.apache.catalina.util.LifecycleSupport.fireLifecycleEvent(LifecycleSupport.java:117) at org.apache.catalina.core.StandardServer.start(StandardServer.java:741) at org.apache.catalina.startup.Catalina.start(Catalina.java:575) This happens because DefaultMCMPHandler$Proxy.getConnection hands the address/port straight to the socket factory and so Socket.connect is called without any timeout. Instead, DefaultSocketFactory.createSocket should be called without a given ip/port so that the returned Socket can then be bound with a proper timeout through Socket.connect(SocketAddress endpoint, int timeout). -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.jboss.org/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira