[ https://issues.apache.org/jira/browse/QPID-8401?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17014413#comment-17014413
]
Robert Godfrey commented on QPID-8401:
--------------------------------------
{quote}Now regarding the current behavior of the broker, note that when the broker dies this
way I end up with a message that stays acquired forever (not by any client consumer but by
the housekeeping task itself). When I restart the broker this message never gets deleted even
though it expired and I can't consume it either... Shouldn't the message be released when
rolling back the failed dequeue operation? [https://github.com/apache/qpid-broker-j/blob/master/broker-core/src/main/java/org/apache/qpid/server/queue/AbstractQueue.java#L1852]
{quote}
So acquisition should not be persisted, so when the broker dies and then restarts the message
should not be acquired. In terms of the onRollback in the code quoted - that logic is for
when a transaction is manually rolled back by a client (not applicable in this case), it is
not designed for handling the case where the broker initiates rollback - in this case the
expected behaviour is that the broker (really should be the virtual host) dies.
> [Broker-J] Broker dies when DB connection is lost
> -------------------------------------------------
>
> Key: QPID-8401
> URL: https://issues.apache.org/jira/browse/QPID-8401
> Project: Qpid
> Issue Type: Bug
> Components: Broker-J
> Affects Versions: qpid-java-broker-7.1.6
> Reporter: Olivier VERMEULEN
> Priority: Critical
>
> When using a JDBC message store, if the housekeeping task is triggered while the DB connection
is lost (DB down or network problem) then the Broker dies with the stack below.
> This happens when a message expires and the housekeeping task tries to delete it from
the store while the DB is not accessible. In this case a StoreException is thrown but this
exception is not catched by the Housekeeping task which is only catching ConnectionScopedRuntimeExceptions.
>
> 2019-12-12 16:22:40,671 ERROR [virtualhost-default-pool-3] (o.a.q.s.Main) - Uncaught
exception, shutting down.
> org.apache.qpid.server.store.StoreException: java.sql.SQLException: JZ006: Caught IOException:
com.sybase.jdbc4.jdbc.SybConnectionDeadException: JZ0C0: Connection is already closed.
> at org.apache.qpid.server.store.jdbc.AbstractJDBCMessageStore$JDBCTransaction.<init>(AbstractJDBCMessageStore.java:1153)
> at org.apache.qpid.server.store.jdbc.GenericAbstractJDBCMessageStore$RecordedJDBCTransaction.<init>(GenericAbstractJDBCMessageStore.java:122)
> at org.apache.qpid.server.store.jdbc.GenericAbstractJDBCMessageStore$RecordedJDBCTransaction.<init>(GenericAbstractJDBCMessageStore.java:118)
> at org.apache.qpid.server.store.jdbc.GenericAbstractJDBCMessageStore.newTransaction(GenericAbstractJDBCMessageStore.java:114)
> at org.apache.qpid.server.txn.AutoCommitTransaction.dequeue(AutoCommitTransaction.java:87)
> at org.apache.qpid.server.queue.AbstractQueue.dequeueEntry(AbstractQueue.java:1780)
> at org.apache.qpid.server.queue.AbstractQueue.dequeueEntry(AbstractQueue.java:1775)
> at org.apache.qpid.server.queue.AbstractQueue.deleteEntry(AbstractQueue.java:1819)
> at org.apache.qpid.server.queue.AbstractQueue.expireEntry(AbstractQueue.java:2354)
> at org.apache.qpid.server.queue.AbstractQueue.getNextAvailableEntry(AbstractQueue.java:2236)
> at org.apache.qpid.server.queue.AbstractQueue.access$1800(AbstractQueue.java:131)
> at org.apache.qpid.server.queue.AbstractQueue$AdvanceConsumersTask.execute(AbstractQueue.java:3712)
> at org.apache.qpid.server.virtualhost.HouseKeepingTask$1.run(HouseKeepingTask.java:56)
> at java.security.AccessController.doPrivileged(Native Method)
> at org.apache.qpid.server.virtualhost.HouseKeepingTask.run(HouseKeepingTask.java:51)
> at java.util.concurrent.Executors$RunnableAdapter.call(Executors.java:511)
> at java.util.concurrent.FutureTask.runAndReset(FutureTask.java:308)
> at java.util.concurrent.ScheduledThreadPoolExecutor$ScheduledFutureTask.access$301(ScheduledThreadPoolExecutor.java:180)
> at java.util.concurrent.ScheduledThreadPoolExecutor$ScheduledFutureTask.run(ScheduledThreadPoolExecutor.java:294)
> at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1142)
> at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:617)
> at org.apache.qpid.server.bytebuffer.QpidByteBufferFactory.lambda$null$0(QpidByteBufferFactory.java:464)
> at java.lang.Thread.run(Thread.java:748)
> Caused by: java.sql.SQLException: JZ006: Caught IOException: com.sybase.jdbc4.jdbc.SybConnectionDeadException:
JZ0C0: Connection is already closed.
> at com.sybase.jdbc4.jdbc.ErrorMessage.createIOEKilledConnEx(ErrorMessage.java:1155)
> at com.sybase.jdbc4.jdbc.ErrorMessage.raiseErrorCheckDead(ErrorMessage.java:1194)
> at com.sybase.jdbc4.tds.Tds.handleIOE(Tds.java:5250)
> at com.sybase.jdbc4.tds.Tds.handleIOE(Tds.java:5195)
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@qpid.apache.org
For additional commands, e-mail: dev-help@qpid.apache.org
|