[
https://issues.apache.org/jira/browse/QPID-7701?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15924391#comment-15924391
]
ASF subversion and git services commented on QPID-7701:
-------------------------------------------------------
Commit 1786914 from [~k-wall] in branch 'java/trunk'
[ https://svn.apache.org/r1786914 ]
QPID-7685: QPID-7701: [BDB] Remove possibility of a lock timeout during message
instance recovery by reading uncommitted data.
Also adheres to the JE Cursor API.
> [BDB HA] Cursor#getNext API misuse
> ----------------------------------
>
> Key: QPID-7701
> URL: https://issues.apache.org/jira/browse/QPID-7701
> Project: Qpid
> Issue Type: Bug
> Components: Java Broker
> Reporter: Keith Wall
> Fix For: 0.30, 0.32, qpid-java-6.0, qpid-java-6.1
>
>
> The documentation for methods {{com.sleepycat.je.Cursor#getNext}} and
> {{com.sleepycat.je.Cursor#getSearchKeyRange}} notes the following:
> {quote}
> In a replicated environment, an explicit transaction must have been
> specified when opening the cursor, unless read-uncommitted isolation is
> specified via the {{CursorConfig}} or {{LockMode}}
> parameter.
> {quote}
> Currently methods such as
> {{org.apache.qpid.server.store.berkeleydb.AbstractBDBMessageStore.BDBMessageStoreReader#visitDistributedTransactions}}
> do not do this.
> The documentation does not state the implications of not following this
> advice.
--
This message was sent by Atlassian JIRA
(v6.3.15#6346)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
