[jira] [Commented] (AMQ-6903) Issue With Master and Slave Mode

["Veaceslav Doina (JIRA)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/AMQ-6903?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16678048#comment-16678048 ] 

Veaceslav Doina commented on AMQ-6903:
--------------------------------------

Hello,

We also trying to configure ActiveMQ 5.15.7 in a cluster using shared storage.

Setup is done on AWS.
 # 2 x EC2 instances
 # 1 x EFS

Initially, all worked fine. But after some period of time and master restart, slave stopped to work.

As was mentioned by [~ChakradharDanireddy], removing data directory solve the issue but we lost all data:

rm -rf /efs/activemq-data/kahadb/

 

 

> Issue With Master and Slave Mode
> --------------------------------
>
>                 Key: AMQ-6903
>                 URL: https://issues.apache.org/jira/browse/AMQ-6903
>             Project: ActiveMQ
>          Issue Type: Bug
>    Affects Versions: 5.15.2
>            Reporter: CHAKRADHAR REDDY
>            Priority: Critical
>
> Hi All,
> We have installed Apache AMQ in master slave mode we could see that slave is not becoming up immediately after master is shutdown and we could see below exception.
>  
> FYI: We have many messages in queues and this slave is not up till we clear data in kahadb location and once we have cleaned data in kahadb location there no messages in queues any longer.
> we have lost all messages from queues is there any option we can get messages from db.log etc.. files from kahadb location
> We have received below transport Errors this is reason we have shutdown master to make slave up
> 2018-01-13 04:33:04,032 | ERROR | Could not accept connection from tcp://IP of server containing interfaces accessing AMQ server:52808 : {} | org.apache.activemq.broker.TransportConnector | ActiveMQ BrokerServ
> ice[ip] Task-15549
> java.lang.IllegalStateException: Timer already cancelled.
> at java.util.Timer.sched(Timer.java:397)[:1.8.0_151]
> at java.util.Timer.schedule(Timer.java:193)[:1.8.0_151]
> at org.apache.activemq.transport.AbstractInactivityMonitor.startConnectCheckTask(AbstractInactivityMonitor.java:425)[activemq-client-5.15.2.jar:5.15.2]
> at org.apache.activemq.transport.AbstractInactivityMonitor.startConne
> After shutdown of master slave is not up with below error and we lost all data in queues.
> 2018-02-21 20:09:18,135 | ERROR | File system space reported by: /DBA/EFS/activemq/kahadb was negative, possibly a huge file system, set a sane usage.total to provide some guidance | org.apache.activemq.broker.BrokerService | main
> 2018-02-21 20:09:18,139 | ERROR | Failed to start Apache ActiveMQ (IP, ID:DNS of activemq server) | org.apache.activemq.broker.BrokerService | main
> org.apache.activemq.ConfigurationException: File system space reported by: /DBA/EFS/activemq/kahadb was negative, possibly a huge file system, set a sane usage.total to provide some guidance
> at org.apache.activemq.broker.BrokerService.checkUsageLimit(BrokerService.java:2092)[activemq-broker-5.15.2.jar:5.15.2]
> at org.apache.activemq.broker.BrokerService.checkStoreUsageLimits(BrokerService.java:2029)[activemq-broker-5.15.2.jar:5.15.2]
> at org.apache.activemq.broker.BrokerService.checkStoreSystemUsageLimits(BrokerService.java:2202)[activemq-broker-5.15.2.jar:5.15.2]
> at org.apache.activemq.broker.BrokerService.doStartBroker(BrokerService.java:777)[activemq-broker-5.15.2.jar:5.15.2]
> at org.apache.activemq.broker.BrokerService.startBroker(BrokerService.java:733)[activemq-broker-5.15.2.jar:5.15.2]
> at org.apache.activemq.broker.BrokerService.start(BrokerService.java:636)[activemq-broker-5.15.2.jar:5.15.2]
> at org.apache.activemq.xbean.XBeanBrokerService.afterPropertiesSet(XBeanBrokerService.java:73)[activemq-spring-5.15.2.jar:5.15.2]
> at sun.reflect.NativeMethodAccessorImpl.invoke0(Native Method)[:1.8.0_151]
> at sun.reflect.NativeMethodAccessorImpl.invoke(NativeMethodAccessorImpl.java:62)[:1.8.0_151]
> :



--
This message was sent by Atlassian JIRA
(v7.6.3#76005)