[jboss-jira] [JBoss JIRA] (AS7-3608) Strange hangup during stopping of server - must be killed by SIGKILL

[Pavel Janousek (JIRA)]

     [ https://issues.jboss.org/browse/AS7-3608?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Pavel Janousek updated AS7-3608:
--------------------------------

    Description: 
During investigation another JIRA (AS7-3607) I've affected by another issue too.

Let say - you have machine after successful reboot. You should prepare network environment as described in AS7-3607, the only one difference is you must disable loopback interface too (ip link set dev lo down).

Now, when you try to start AS7, everything is OK, more precisely is the same as in other issue (AS7-3607).

BUT - when you hit Ctrl-C now - you try to stop server - it hang-up and report only:
{code}
17:05:23,502 INFO  [org.jboss.as] (Controller Boot Thread) JBoss EAP 6.0.0.Alpha2 (AS 7.1.0.CR1-redhat-1) started in 14012ms - Started 158 of 228 services (68 services are passive or on-demand)
^C17:05:50,274 INFO  [org.jboss.as.messaging] (MSC service thread 1-2) JBAS011605: Unbound messaging object to jndi name java:/queue/test
17:05:50,279 INFO  [org.jboss.as.osgi] (MSC service thread 1-1) JBAS011921: Stopping OSGi Framework
17:05:50,279 INFO  [org.jboss.as.messaging] (MSC service thread 1-2) JBAS011605: Unbound messaging object to jndi name java:/RemoteConnectionFactory
17:05:50,282 INFO  [org.jboss.as.messaging] (MSC service thread 1-2) JBAS011605: Unbound messaging object to jndi name java:/ConnectionFactory
17:05:50,341 INFO  [org.jboss.as.logging] JBAS011503: Restored bootstrap log handlers
17:05:50,350 INFO  [jacorb.poa] POA IRPOA destroyed
17:05:50,381 INFO  [org.hornetq.core.server.impl.HornetQServerImpl] HornetQ Server version 2.2.7.Final (HQ_2_2_7_FINAL_AS7, 121) [3fe9aa6a-4d86-11e1-b2c6-000000000001] stopped
17:05:50,411 INFO  [jacorb.poa] POA Naming destroyed
17:05:50,434 INFO  [jacorb.poa] POA RootPOA destroyed
17:05:50,440 INFO  [com.arjuna.ats.jbossatx] ARJUNA032018: Destroying TransactionManagerService
17:05:50,442 INFO  [com.arjuna.ats.jbossatx] ARJUNA032014: Stopping transaction recovery manager
17:05:50,441 INFO  [jacorb.orb] prepare ORB for shutdown...
17:05:50,442 INFO  [jacorb.orb] ORB going down...
17:05:50,446 INFO  [jacorb.orb.iiop] Listener exited
17:05:50,447 INFO  [jacorb.orb] ORB shutdown complete
17:05:50,447 INFO  [jacorb.orb] ORB run, exit
^Z
[1]+  Stopped                 ./standalone.sh
[pjanouse at fedora15-vrt1 bin]$ kill -9 %1
[1]+  Killed                  ./standalone.sh
[pjanouse at fedora15-vrt1 bin]$
{code}

As you can see, it can be killed only by SIGKILL signal.

More strange is when you now enable loopback interface again (ip link set dev lo up), start/stop sequence works fine (as expected).

And the strangest is when you disable loopback again, it works fine as in previous step - everything works as you can expect. I think it is related to some network issue, but I can't be more precise now, I can't find any other aspect why it is occurred after reboot only.

I try also several attempts (loopback interface disabled) during let say 10 minutes (due to attempt to eliminate resolver/DNS timeouts etc.), but until I did up/down sequence in lo interface, behavior was the same - only SIGKILL stops it.

I try it several times, after several reboot my instance inside VirtualBox and behavior is the same - so it isn't accind situation. So I hope it isn't related to only network stack implementation/virtualization inside VirtualBox and is reproducible in another linux box too.

  was:
During investigation another JIRA (JBPAPP-8045) I've affected by another issue too.

Let say - you have machine after successful reboot. You should prepare network environment as described in JBPAPP-8045, the only one difference is you must disable loopback interface too (ip link set dev lo down).

Now, when you try to start AS7, everything is OK, more precisely is the same as in other issue (JBPAPP-8045).

BUT - when you hit Ctrl-C now - you try to stop server - it hang-up and report only:
{code}
17:05:23,502 INFO  [org.jboss.as] (Controller Boot Thread) JBoss EAP 6.0.0.Alpha2 (AS 7.1.0.CR1-redhat-1) started in 14012ms - Started 158 of 228 services (68 services are passive or on-demand)
^C17:05:50,274 INFO  [org.jboss.as.messaging] (MSC service thread 1-2) JBAS011605: Unbound messaging object to jndi name java:/queue/test
17:05:50,279 INFO  [org.jboss.as.osgi] (MSC service thread 1-1) JBAS011921: Stopping OSGi Framework
17:05:50,279 INFO  [org.jboss.as.messaging] (MSC service thread 1-2) JBAS011605: Unbound messaging object to jndi name java:/RemoteConnectionFactory
17:05:50,282 INFO  [org.jboss.as.messaging] (MSC service thread 1-2) JBAS011605: Unbound messaging object to jndi name java:/ConnectionFactory
17:05:50,341 INFO  [org.jboss.as.logging] JBAS011503: Restored bootstrap log handlers
17:05:50,350 INFO  [jacorb.poa] POA IRPOA destroyed
17:05:50,381 INFO  [org.hornetq.core.server.impl.HornetQServerImpl] HornetQ Server version 2.2.7.Final (HQ_2_2_7_FINAL_AS7, 121) [3fe9aa6a-4d86-11e1-b2c6-000000000001] stopped
17:05:50,411 INFO  [jacorb.poa] POA Naming destroyed
17:05:50,434 INFO  [jacorb.poa] POA RootPOA destroyed
17:05:50,440 INFO  [com.arjuna.ats.jbossatx] ARJUNA032018: Destroying TransactionManagerService
17:05:50,442 INFO  [com.arjuna.ats.jbossatx] ARJUNA032014: Stopping transaction recovery manager
17:05:50,441 INFO  [jacorb.orb] prepare ORB for shutdown...
17:05:50,442 INFO  [jacorb.orb] ORB going down...
17:05:50,446 INFO  [jacorb.orb.iiop] Listener exited
17:05:50,447 INFO  [jacorb.orb] ORB shutdown complete
17:05:50,447 INFO  [jacorb.orb] ORB run, exit
^Z
[1]+  Stopped                 ./standalone.sh
[pjanouse at fedora15-vrt1 bin]$ kill -9 %1
[1]+  Killed                  ./standalone.sh
[pjanouse at fedora15-vrt1 bin]$
{code}

As you can see, it can be killed only by SIGKILL signal.

More strange is when you now enable loopback interface again (ip link set dev lo up), start/stop sequence works fine (as expected).

And the strangest is when you disable loopback again, it works fine as in previous step - everything works as you can expect. I think it is related to some network issue, but I can't be more precise now, I can't find any other aspect why it is occurred after reboot only.

I try also several attempts (loopback interface disabled) during let say 10 minutes (due to attempt to eliminate resolver/DNS timeouts etc.), but until I did up/down sequence in lo interface, behavior was the same - only SIGKILL stops it.

I try it several times, after several reboot my instance inside VirtualBox and behavior is the same - so it isn't accind situation. So I hope it isn't related to only network stack implementation/virtualization inside VirtualBox and is reproducible in another linux box too.


    
> Strange hangup during stopping of server - must be killed by SIGKILL
> --------------------------------------------------------------------
>
>                 Key: AS7-3608
>                 URL: https://issues.jboss.org/browse/AS7-3608
>             Project: Application Server 7
>          Issue Type: Bug
>    Affects Versions: 7.1.0.CR1b
>            Reporter: Pavel Janousek
>            Priority: Critical
>
> During investigation another JIRA (AS7-3607) I've affected by another issue too.
> Let say - you have machine after successful reboot. You should prepare network environment as described in AS7-3607, the only one difference is you must disable loopback interface too (ip link set dev lo down).
> Now, when you try to start AS7, everything is OK, more precisely is the same as in other issue (AS7-3607).
> BUT - when you hit Ctrl-C now - you try to stop server - it hang-up and report only:
> {code}
> 17:05:23,502 INFO  [org.jboss.as] (Controller Boot Thread) JBoss EAP 6.0.0.Alpha2 (AS 7.1.0.CR1-redhat-1) started in 14012ms - Started 158 of 228 services (68 services are passive or on-demand)
> ^C17:05:50,274 INFO  [org.jboss.as.messaging] (MSC service thread 1-2) JBAS011605: Unbound messaging object to jndi name java:/queue/test
> 17:05:50,279 INFO  [org.jboss.as.osgi] (MSC service thread 1-1) JBAS011921: Stopping OSGi Framework
> 17:05:50,279 INFO  [org.jboss.as.messaging] (MSC service thread 1-2) JBAS011605: Unbound messaging object to jndi name java:/RemoteConnectionFactory
> 17:05:50,282 INFO  [org.jboss.as.messaging] (MSC service thread 1-2) JBAS011605: Unbound messaging object to jndi name java:/ConnectionFactory
> 17:05:50,341 INFO  [org.jboss.as.logging] JBAS011503: Restored bootstrap log handlers
> 17:05:50,350 INFO  [jacorb.poa] POA IRPOA destroyed
> 17:05:50,381 INFO  [org.hornetq.core.server.impl.HornetQServerImpl] HornetQ Server version 2.2.7.Final (HQ_2_2_7_FINAL_AS7, 121) [3fe9aa6a-4d86-11e1-b2c6-000000000001] stopped
> 17:05:50,411 INFO  [jacorb.poa] POA Naming destroyed
> 17:05:50,434 INFO  [jacorb.poa] POA RootPOA destroyed
> 17:05:50,440 INFO  [com.arjuna.ats.jbossatx] ARJUNA032018: Destroying TransactionManagerService
> 17:05:50,442 INFO  [com.arjuna.ats.jbossatx] ARJUNA032014: Stopping transaction recovery manager
> 17:05:50,441 INFO  [jacorb.orb] prepare ORB for shutdown...
> 17:05:50,442 INFO  [jacorb.orb] ORB going down...
> 17:05:50,446 INFO  [jacorb.orb.iiop] Listener exited
> 17:05:50,447 INFO  [jacorb.orb] ORB shutdown complete
> 17:05:50,447 INFO  [jacorb.orb] ORB run, exit
> ^Z
> [1]+  Stopped                 ./standalone.sh
> [pjanouse at fedora15-vrt1 bin]$ kill -9 %1
> [1]+  Killed                  ./standalone.sh
> [pjanouse at fedora15-vrt1 bin]$
> {code}
> As you can see, it can be killed only by SIGKILL signal.
> More strange is when you now enable loopback interface again (ip link set dev lo up), start/stop sequence works fine (as expected).
> And the strangest is when you disable loopback again, it works fine as in previous step - everything works as you can expect. I think it is related to some network issue, but I can't be more precise now, I can't find any other aspect why it is occurred after reboot only.
> I try also several attempts (loopback interface disabled) during let say 10 minutes (due to attempt to eliminate resolver/DNS timeouts etc.), but until I did up/down sequence in lo interface, behavior was the same - only SIGKILL stops it.
> I try it several times, after several reboot my instance inside VirtualBox and behavior is the same - so it isn't accind situation. So I hope it isn't related to only network stack implementation/virtualization inside VirtualBox and is reproducible in another linux box too.

--
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