curator-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Jordan Zimmerman (JIRA)" <j...@apache.org>
Subject [jira] [Created] (CURATOR-24) The current method of managing hung ZK handles needs improvement
Date Fri, 10 May 2013 00:07:15 GMT
Jordan Zimmerman created CURATOR-24:
---------------------------------------

             Summary: The current method of managing hung ZK handles needs improvement
                 Key: CURATOR-24
                 URL: https://issues.apache.org/jira/browse/CURATOR-24
             Project: Apache Curator
          Issue Type: Improvement
          Components: Client
    Affects Versions: 2.0.0-incubating
            Reporter: Jordan Zimmerman
            Assignee: Jordan Zimmerman
             Fix For: 2.0.1-incubating


In v1.3.0, a "major change" was added whereby "when the Curator state changes to LOST, a flag
will be set
so that the next time Curator needs to get the ZooKeeper instance, the current instance will
be closed and a new ZooKeeper instance will be allocated (as if the session had expired)."

This has turned out not to be optimum. Instead, if the session timeout elapses before a SysConnected
is received, treat it as a failed session and dispose and reallocate the ZooKeeper handle.
This has be shown to be superior internally at Netflix.

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira

Mime
View raw message