Re: Hang on shutdown on Harmony
I'm not sure I understand how this stack trace is occurring. Is it possible that there's an error in Harmony exception handling? Here's what's supposed to be happening here. At ORB shutdown, the socket that acceptor has issued the accept() call is closed. That's line 111 in Acceptor_impl. This should cause an I/O exception, which is what the stack trace is showing. However, there's a catch for that IOException at line 125 in Acceptor_impl that should be turning that into a org.omg.CORBA.COMM_FAILURE exception. That exception in turn should be caught GIOPServerStartedThreaded.starterRun(), which detects at that point that this is a shutdown situation and terminates the run loop on the thread. The loop that Jarek mentioned in ORBControl is waiting for the server threads to finish shutting down, but if that IOException is not getting trapped properly, this thread goes away without cleanuing up its state. I am a little puzzled about how completeServerShutdown can get caught in a loop, since it appears to have a cap on how many times it will go around that loop. Rick Zakharov, Vasily M wrote: Hi, all, I observe the following stack trace when shutting down Geronimo (branches/2.0) running on Harmony, and then Geronimo stops responding forever. Otherwise, Geronimo runs pretty well on Harmony, see [1], details are at [2]. What is interesting is exactly similar stack and hang are observed when running org.apache.geronimo.yoko.TSSConfigEditorTest test on Harmony, I already wrote about it, see [3]. For now I couldn't find out the cause for this behavior, so if someone has any ideas on where to look, I'd be grateful. Thanks! [] received stop signal Jan 24, 2008 5:34:18 PM org.apache.yoko.orb.OCI.IIOP.Acceptor_impl accept SEVERE: Failure accepting connection for host=ecwm0vmzakhar/192.168.16.2, port=2001 Throwable occurred: java.io.IOException: Connection was closed at org.apache.harmony.xnet.provider.jsse.SSLSocketImpl.doHandshake(SSLSocke tImpl.java:760) at org.apache.harmony.xnet.provider.jsse.SSLSocketImpl.startHandshake(SSLSo cketImpl.java:438) at org.apache.harmony.xnet.provider.jsse.SSLServerSocketImpl.accept(SSLServ erSocketImpl.java:256) at org.apache.yoko.orb.OCI.IIOP.Acceptor_impl.accept(Acceptor_impl.java:111 ) at org.apache.yoko.orb.OB.GIOPServerStarterThreaded.starterRun(GIOPServerSt arterThreaded.java:195) at org.apache.yoko.orb.OB.GIOPServerStarterThreaded$StarterThread.run(GIOPS erverStarterThreaded.java:34) [1] http://thread.gmane.org/gmane.comp.java.geronimo.devel/57921 [2] http://cwiki.apache.org/confluence/display/GMOxDOC20/Apache+Harmony [3] http://thread.gmane.org/gmane.comp.java.geronimo.devel/57396 Vasily Zakharov, Intel ESSD --- Closed Joint Stock Company Intel A/O Registered legal address: 125252, Moscow, Russian Federation, Chapayevsky Per, 14. This e-mail and any attachments may contain confidential material for the sole use of the intended recipient(s). Any review or distribution by others is strictly prohibited. If you are not the intended recipient, please contact the sender and delete all copies.
RE: Hang on shutdown on Harmony
Rick, Thanks for the detailed analysys, I'll investigate it. Vasily -Original Message- From: Rick McGuire [mailto:[EMAIL PROTECTED] Sent: Friday, January 25, 2008 1:56 PM To: dev@geronimo.apache.org Subject: Re: Hang on shutdown on Harmony I'm not sure I understand how this stack trace is occurring. Is it possible that there's an error in Harmony exception handling? Here's what's supposed to be happening here. At ORB shutdown, the socket that acceptor has issued the accept() call is closed. That's line 111 in Acceptor_impl. This should cause an I/O exception, which is what the stack trace is showing. However, there's a catch for that IOException at line 125 in Acceptor_impl that should be turning that into a org.omg.CORBA.COMM_FAILURE exception. That exception in turn should be caught GIOPServerStartedThreaded.starterRun(), which detects at that point that this is a shutdown situation and terminates the run loop on the thread. The loop that Jarek mentioned in ORBControl is waiting for the server threads to finish shutting down, but if that IOException is not getting trapped properly, this thread goes away without cleanuing up its state. I am a little puzzled about how completeServerShutdown can get caught in a loop, since it appears to have a cap on how many times it will go around that loop. Rick Zakharov, Vasily M wrote: Hi, all, I observe the following stack trace when shutting down Geronimo (branches/2.0) running on Harmony, and then Geronimo stops responding forever. Otherwise, Geronimo runs pretty well on Harmony, see [1], details are at [2]. What is interesting is exactly similar stack and hang are observed when running org.apache.geronimo.yoko.TSSConfigEditorTest test on Harmony, I already wrote about it, see [3]. For now I couldn't find out the cause for this behavior, so if someone has any ideas on where to look, I'd be grateful. Thanks! [] received stop signal Jan 24, 2008 5:34:18 PM org.apache.yoko.orb.OCI.IIOP.Acceptor_impl accept SEVERE: Failure accepting connection for host=ecwm0vmzakhar/192.168.16.2, port=2001 Throwable occurred: java.io.IOException: Connection was closed at org.apache.harmony.xnet.provider.jsse.SSLSocketImpl.doHandshake(SSLSocke tImpl.java:760) at org.apache.harmony.xnet.provider.jsse.SSLSocketImpl.startHandshake(SSLSo cketImpl.java:438) at org.apache.harmony.xnet.provider.jsse.SSLServerSocketImpl.accept(SSLServ erSocketImpl.java:256) at org.apache.yoko.orb.OCI.IIOP.Acceptor_impl.accept(Acceptor_impl.java:111 ) at org.apache.yoko.orb.OB.GIOPServerStarterThreaded.starterRun(GIOPServerSt arterThreaded.java:195) at org.apache.yoko.orb.OB.GIOPServerStarterThreaded$StarterThread.run(GIOPS erverStarterThreaded.java:34) [1] http://thread.gmane.org/gmane.comp.java.geronimo.devel/57921 [2] http://cwiki.apache.org/confluence/display/GMOxDOC20/Apache+Harmony [3] http://thread.gmane.org/gmane.comp.java.geronimo.devel/57396 Vasily Zakharov, Intel ESSD --- Closed Joint Stock Company Intel A/O Registered legal address: 125252, Moscow, Russian Federation, Chapayevsky Per, 14. This e-mail and any attachments may contain confidential material for the sole use of the intended recipient(s). Any review or distribution by others is strictly prohibited. If you are not the intended recipient, please contact the sender and delete all copies. Closed Joint Stock Company Intel A/O Registered legal address: 125252, Moscow, Russian Federation, Chapayevsky Per, 14. This e-mail and any attachments may contain confidential material for the sole use of the intended recipient(s). Any review or distribution by others is strictly prohibited. If you are not the intended recipient, please contact the sender and delete all copies.
Hang on shutdown on Harmony
Hi, all, I observe the following stack trace when shutting down Geronimo (branches/2.0) running on Harmony, and then Geronimo stops responding forever. Otherwise, Geronimo runs pretty well on Harmony, see [1], details are at [2]. What is interesting is exactly similar stack and hang are observed when running org.apache.geronimo.yoko.TSSConfigEditorTest test on Harmony, I already wrote about it, see [3]. For now I couldn't find out the cause for this behavior, so if someone has any ideas on where to look, I'd be grateful. Thanks! [] received stop signal Jan 24, 2008 5:34:18 PM org.apache.yoko.orb.OCI.IIOP.Acceptor_impl accept SEVERE: Failure accepting connection for host=ecwm0vmzakhar/192.168.16.2, port=2001 Throwable occurred: java.io.IOException: Connection was closed at org.apache.harmony.xnet.provider.jsse.SSLSocketImpl.doHandshake(SSLSocke tImpl.java:760) at org.apache.harmony.xnet.provider.jsse.SSLSocketImpl.startHandshake(SSLSo cketImpl.java:438) at org.apache.harmony.xnet.provider.jsse.SSLServerSocketImpl.accept(SSLServ erSocketImpl.java:256) at org.apache.yoko.orb.OCI.IIOP.Acceptor_impl.accept(Acceptor_impl.java:111 ) at org.apache.yoko.orb.OB.GIOPServerStarterThreaded.starterRun(GIOPServerSt arterThreaded.java:195) at org.apache.yoko.orb.OB.GIOPServerStarterThreaded$StarterThread.run(GIOPS erverStarterThreaded.java:34) [1] http://thread.gmane.org/gmane.comp.java.geronimo.devel/57921 [2] http://cwiki.apache.org/confluence/display/GMOxDOC20/Apache+Harmony [3] http://thread.gmane.org/gmane.comp.java.geronimo.devel/57396 Vasily Zakharov, Intel ESSD --- Closed Joint Stock Company Intel A/O Registered legal address: 125252, Moscow, Russian Federation, Chapayevsky Per, 14. This e-mail and any attachments may contain confidential material for the sole use of the intended recipient(s). Any review or distribution by others is strictly prohibited. If you are not the intended recipient, please contact the sender and delete all copies.
Re: Hang on shutdown on Harmony
I'm trying to remember but once I managed to get Harmony running in a debugger (while debugging the TSSConfigEditorTest test issue) I've noticed that it was stuck in a loop in ORBControl.completeServerShutdown() (http://svn.apache.org/viewvc/geronimo/yoko/trunk/core/src/main/java/org/apache/yoko/orb/OB/ORBControl.java?view=markup) Jarek On Jan 24, 2008 11:19 AM, Zakharov, Vasily M [EMAIL PROTECTED] wrote: Hi, all, I observe the following stack trace when shutting down Geronimo (branches/2.0) running on Harmony, and then Geronimo stops responding forever. Otherwise, Geronimo runs pretty well on Harmony, see [1], details are at [2]. What is interesting is exactly similar stack and hang are observed when running org.apache.geronimo.yoko.TSSConfigEditorTest test on Harmony, I already wrote about it, see [3]. For now I couldn't find out the cause for this behavior, so if someone has any ideas on where to look, I'd be grateful. Thanks! [] received stop signal Jan 24, 2008 5:34:18 PM org.apache.yoko.orb.OCI.IIOP.Acceptor_impl accept SEVERE: Failure accepting connection for host=ecwm0vmzakhar/192.168.16.2, port=2001 Throwable occurred: java.io.IOException: Connection was closed at org.apache.harmony.xnet.provider.jsse.SSLSocketImpl.doHandshake(SSLSocke tImpl.java:760) at org.apache.harmony.xnet.provider.jsse.SSLSocketImpl.startHandshake(SSLSo cketImpl.java:438) at org.apache.harmony.xnet.provider.jsse.SSLServerSocketImpl.accept(SSLServ erSocketImpl.java:256) at org.apache.yoko.orb.OCI.IIOP.Acceptor_impl.accept(Acceptor_impl.java:111 ) at org.apache.yoko.orb.OB.GIOPServerStarterThreaded.starterRun(GIOPServerSt arterThreaded.java:195) at org.apache.yoko.orb.OB.GIOPServerStarterThreaded$StarterThread.run(GIOPS erverStarterThreaded.java:34) [1] http://thread.gmane.org/gmane.comp.java.geronimo.devel/57921 [2] http://cwiki.apache.org/confluence/display/GMOxDOC20/Apache+Harmony [3] http://thread.gmane.org/gmane.comp.java.geronimo.devel/57396 Vasily Zakharov, Intel ESSD --- Closed Joint Stock Company Intel A/O Registered legal address: 125252, Moscow, Russian Federation, Chapayevsky Per, 14. This e-mail and any attachments may contain confidential material for the sole use of the intended recipient(s). Any review or distribution by others is strictly prohibited. If you are not the intended recipient, please contact the sender and delete all copies.
RE: Hang on shutdown on Harmony
Jarek, Thanks! I'll check that. Vasily -Original Message- From: Jarek Gawor [mailto:[EMAIL PROTECTED] Sent: Thursday, January 24, 2008 9:24 PM To: dev@geronimo.apache.org Subject: Re: Hang on shutdown on Harmony I'm trying to remember but once I managed to get Harmony running in a debugger (while debugging the TSSConfigEditorTest test issue) I've noticed that it was stuck in a loop in ORBControl.completeServerShutdown() (http://svn.apache.org/viewvc/geronimo/yoko/trunk/core/src/main/java/org /apache/yoko/orb/OB/ORBControl.java?view=markup) Jarek On Jan 24, 2008 11:19 AM, Zakharov, Vasily M [EMAIL PROTECTED] wrote: Hi, all, I observe the following stack trace when shutting down Geronimo (branches/2.0) running on Harmony, and then Geronimo stops responding forever. Otherwise, Geronimo runs pretty well on Harmony, see [1], details are at [2]. What is interesting is exactly similar stack and hang are observed when running org.apache.geronimo.yoko.TSSConfigEditorTest test on Harmony, I already wrote about it, see [3]. For now I couldn't find out the cause for this behavior, so if someone has any ideas on where to look, I'd be grateful. Thanks! [] received stop signal Jan 24, 2008 5:34:18 PM org.apache.yoko.orb.OCI.IIOP.Acceptor_impl accept SEVERE: Failure accepting connection for host=ecwm0vmzakhar/192.168.16.2, port=2001 Throwable occurred: java.io.IOException: Connection was closed at org.apache.harmony.xnet.provider.jsse.SSLSocketImpl.doHandshake(SSLSocke tImpl.java:760) at org.apache.harmony.xnet.provider.jsse.SSLSocketImpl.startHandshake(SSLSo cketImpl.java:438) at org.apache.harmony.xnet.provider.jsse.SSLServerSocketImpl.accept(SSLServ erSocketImpl.java:256) at org.apache.yoko.orb.OCI.IIOP.Acceptor_impl.accept(Acceptor_impl.java:111 ) at org.apache.yoko.orb.OB.GIOPServerStarterThreaded.starterRun(GIOPServerSt arterThreaded.java:195) at org.apache.yoko.orb.OB.GIOPServerStarterThreaded$StarterThread.run(GIOPS erverStarterThreaded.java:34) [1] http://thread.gmane.org/gmane.comp.java.geronimo.devel/57921 [2] http://cwiki.apache.org/confluence/display/GMOxDOC20/Apache+Harmony [3] http://thread.gmane.org/gmane.comp.java.geronimo.devel/57396 Vasily Zakharov, Intel ESSD --- Closed Joint Stock Company Intel A/O Registered legal address: 125252, Moscow, Russian Federation, Chapayevsky Per, 14. This e-mail and any attachments may contain confidential material for the sole use of the intended recipient(s). Any review or distribution by others is strictly prohibited. If you are not the intended recipient, please contact the sender and delete all copies. Closed Joint Stock Company Intel A/O Registered legal address: 125252, Moscow, Russian Federation, Chapayevsky Per, 14. This e-mail and any attachments may contain confidential material for the sole use of the intended recipient(s). Any review or distribution by others is strictly prohibited. If you are not the intended recipient, please contact the sender and delete all copies.
