You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@cassandra.apache.org by "Philip Thompson (JIRA)" <ji...@apache.org> on 2015/05/18 22:42:00 UTC

[jira] [Commented] (CASSANDRA-9419) CqlRecordWriter does not adequately determine if driver failed to connect due to interrupt

    [ https://issues.apache.org/jira/browse/CASSANDRA-9419?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14549183#comment-14549183 ] 

Philip Thompson commented on CASSANDRA-9419:
--------------------------------------------

Though the previous method did seem to be working reliably in testing, changes to trunk since have caused intermittent failures where the interrupt flag is cleared, and we don't retry connections.

> CqlRecordWriter does not adequately determine if driver failed to connect due to interrupt
> ------------------------------------------------------------------------------------------
>
>                 Key: CASSANDRA-9419
>                 URL: https://issues.apache.org/jira/browse/CASSANDRA-9419
>             Project: Cassandra
>          Issue Type: Bug
>            Reporter: Philip Thompson
>            Assignee: Philip Thompson
>             Fix For: 2.2 rc1
>
>         Attachments: 9419.txt
>
>
> Until https://datastax-oss.atlassian.net/browse/JAVA-769 is implemented, I have had to implement a workaround.
> Java's thread interrupt status flag can be cleared for any number of reasons. After discussion, I've changed to parsing the driver's DriverException to tell if we are safe to retry the failed connection.



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)