cassandra-commits mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Jonathan Ellis (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (CASSANDRA-4053) IncomingTcpConnection can not be closed when the peer is brutaly terminated or switch is failed
Date Mon, 23 Sep 2013 13:22:22 GMT

    [ https://issues.apache.org/jira/browse/CASSANDRA-4053?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13774529#comment-13774529
] 

Jonathan Ellis commented on CASSANDRA-4053:
-------------------------------------------

+1
                
> IncomingTcpConnection can not be closed when the peer is brutaly terminated or switch
is failed
> -----------------------------------------------------------------------------------------------
>
>                 Key: CASSANDRA-4053
>                 URL: https://issues.apache.org/jira/browse/CASSANDRA-4053
>             Project: Cassandra
>          Issue Type: Bug
>          Components: Core
>    Affects Versions: 1.0.8
>            Reporter: Zhu Han
>            Assignee: Marcus Eriksson
>         Attachments: 0001-enable-keepalive-on-incoming-connections.patch
>
>
> IncomingTcpConnection has no way to detect the peer is down when the peer meets power
loss or the network infrastructure is failed, and the thread is leaked...
> For safety, as least SO_KEEPALIVE should be set on those IncomingTcpConnections. The
better way is to close the incoming connections when failure detector notifies the peer failure,
but it requires some extra bookmarking.
> Besides it, it would be better if IncomingTcpConnection and OutgoingTcpConnection is
marked as daemon thread...

--
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