I wrote this page after successfully using DCC on a cluster with 5 OFBiz
servers and 3 Postgres machines with Pgpool-II (1 master, 2 replicants)
This was based on roughly R11.04 (actually a revision taken just before in
February 2011 and then manually updated during 2,5 years for main bugs)
I did not have to use DCC since, but I believe some are using it, the same way I wrote then there, with newer versions. Pierre Smits for instance let
a comment about using ServiceMix rather than AMQ (actually ServiceMix embeds AMQ)
It could be though that changes in newer versions have not been totally checked. Before trying with an older version (if it makes sense to you) did
you check your logs for "cache clear" messages?
HTH
Happy new year to everybody!
Jacques
Le 31/12/2015 17:58, Len Shein a écrit :
All,
We have two instances of Ofbiz running. Both instances talk to the same database. When we update some entity using one of the instances we are
looking for cache clearing operation on all the instances to happen automatically. Based on Ofbiz documentation found on “Distriubted Cache
Clearing” we have setup the following.
Instance One we called the “master”.
·Instance Id: master
·Pool: masterPool
Instance Two we called the “slave”
·Instance Id: slave01
·Pool: slave01Pool
Installed ActiveMq version 5.10.0 on the Master Instance.
On both the Master and Slave Instances the following:
1.In Framework/base/lib/
Copied the ‘activemq-all-5.10.0.jar
2.In Framework/entity/config/entityengine.xml
<delegator name="default" entity-model-reader="main" entity-group-reader="main"
entity-eca-reader="main" distributed-cache-clear-enabled="*true*">
3.In Framework/service/config/serviceengine.xml
<jms-service name="serviceMessenger"send-mode="all">
<server jndi-server-name="*default*"
jndi-name="topicConnectionFactory"
topic-queue="*OFBTopic*"
type="topic"
listen="true"/>
</jms-service>
4.In Framework/base/config/jndi.properties
java.naming.factory.initial=org.apache.activemq.jndi.ActiveMQInitialContextFactory
java.naming.provider.url=tcp://<AMQ-IP-SERVER1>:61616
topic.OFBTopic=*OFBTopic*
connectionFactoryNames=connectionFactory, queueConnectionFactory,
topicConnectionFactory
*Commented out the default jndi.properties*
java.naming.factory.initial=com.sun.jndi.rmi.registry.RegistryContextFactory
java.naming.provider.url=_rmi_://127.0.0.1:1099
With this configuration both instances correctly send messages to the ‘*OFBTopic’. *This is verified by viewing the ActiveMQ admin monitor.
http://<AMQ-IP-SERVER1>:8161/admin/topics.jsp <http://%3cAMQ-IP-SERVER1%3e:8161/admin/topics.jsp>
*However when updating for example the PRODUCT entity on the ‘Master’ instance the ‘cache’ for the Product entity on the ‘Slave’ instance is NOT
cleared.*
*Has anyone succesfully configured the Ditributed Cache clear using ActiveMQ in
the past?*
We have also tried a second configuration
1.In Framework/base/lib/
Copied the ‘activemq-all-5.10.0.jar
2.In Framework/entity/config/entityengine.xml
<delegator name="default" entity-model-reader="main" entity-group-reader="main"
entity-eca-reader="main" distributed-cache-clear-enabled="*true*">
3.In Framework/service/config/serviceengine.xml
<jms-service name="serviceMessenger"send-mode="all">
<server jndi-server-name="*activemq*"
jndi-name="ConnectionFactory"
topic-queue="*OFBTopic*"
type="topic"
listen="true"/>
</jms-service>
4.In Framework/base/config/jndiservers.xml
<jndi-server name="activemq">
Context-provider-url="tcp://<AMQ-IP-SERVER1>:61616?jms.useAsyncSend=true&timeout=6000"
Initial-context-factory ="
org.apache.activemq.jndi.ActiveMQInitialContextFactory "
url-pkg-prefixes=""
security-principal=””
security-credentials=””
/>
5.In Framework/base/config/jndi.properties
java.naming.factory.initial=com.sun.jndi.rmi.registry.RegistryContextFactory
java.naming.provider.url=_rmi_://127.0.0.1:1099
topic.OFBTopic=*OFBTopic*
When using the second type configuration on ‘startup’ we receive the following
erro message in the log and no updates can be performed on the DB.
Failure in storeByCondition operation for entity [JobSandbox] - SQL Exception occurred on commit (Commit can not be set while enrolled in a
transaction)
Error in polling JobSandbox: [JobManager.java:188:ERROR] – SQL Exception
occurred on commit (Commit can not be set while enrolled in a transaction)
*Any sugestions from the community would be of great help?*
Len Shein
lsh...@solveda.com <mailto:lsh...@solveda.com>
Office: 516.742.7888 ext.225
Home Office: 732.333.4303
Cell: 917.882.8515
