activemq-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Martin Serrano (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (AMQ-4164) In network of master slave brokers, if both masters down, slaves incorrectly create treat each other as slaves
Date Wed, 07 Nov 2012 16:23:13 GMT

    [ https://issues.apache.org/jira/browse/AMQ-4164?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13492470#comment-13492470
] 

Martin Serrano commented on AMQ-4164:
-------------------------------------

I believe this is part of what is going [here|http://activemq.2283324.n4.nabble.com/master-slave-within-a-network-of-brokers-td4410862.html]
and in my own [post|http://activemq.2283324.n4.nabble.com/in-master-slave-broker-network-networked-slave-incorrectly-bound-to-other-slave-tt4658858.html]

I think having the incorrect slave connection results in these messages, taken from that first
post:

{noformat}
2012-02-22 10:08:44,763 [ActiveMQ Task-5] WARN  org.apache.activemq.network.DemandForwardingBridgeSupport
 - Unexpected local command: SessionInfo {commandId = 5, ...
2012-02-22 10:08:44,766 [ActiveMQ Task-5] WARN  org.apache.activemq.network.DemandForwardingBridgeSupport
 - Unexpected local command: ProducerInfo {commandId = 6, ...
2012-02-22 10:08:44,768 [ActiveMQ Task-5] WARN  org.apache.activemq.network.DemandForwardingBridgeSupport
 - Unexpected local command: ActiveMQTextMessage {commandId = 7, ...
{noformat}

                
> In network of master slave brokers, if both masters down, slaves incorrectly create treat
each other as slaves
> --------------------------------------------------------------------------------------------------------------
>
>                 Key: AMQ-4164
>                 URL: https://issues.apache.org/jira/browse/AMQ-4164
>             Project: ActiveMQ
>          Issue Type: Bug
>          Components: Transport
>    Affects Versions: 5.7.0
>            Reporter: Martin Serrano
>
> When running a test with a network of master/slave brokers, I have both masters go down.
 In this case, the slaves are started and their respective network connections are started.
 The logic in {{TransportConnection.processBrokerInfo(BrokerInfo info)}}, checks if the info
represents a slave broker and starts a MasterBroker filter if so.  However this logic is faulty
if the info for the slave broker is arriving via a network connection.
> As a result a slave tries to connect a networked broker as a slave.  This results in
incorrect behavior, errors, and can cause the system to lockup during shutdown (due to hanging
ResponseCorrelation).

--
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