cluster cache does not update nodes on object remove

[svkap]

Hi.

I am using Jboss Cache Core 3.2.1 and the last version of Pojo cache. I am running it in cluster mode but there is 1 problem. When an object is removed from 1 of the nodes it is not removed from the other. I have tied using clustering mode="invalidation" but strangely enough the new created objects are not sent to the other node. Most likely all these problems are a matter of configuration. Can you tell me where is my mistake?

<?xml version="1.0" encoding="UTF-8"?>
<jbosscache xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
    xmlns="urn:jboss:jbosscache-core:config:3.2">

    <locking isolationLevel="REPEATABLE_READ"
        lockParentForChildInsertRemove="false" lockAcquisitionTimeout="20000"
        nodeLockingScheme="mvcc" writeSkewCheck="false" useLockStriping="true"
        concurrencyLevel="500" />

    <jmxStatistics enabled="true" />

    <startup regionsInactiveOnStartup="false" />

    <shutdown hookBehavior="DEFAULT" />

    <listeners asyncPoolSize="1" asyncQueueSize="100000" />

    <invocationBatching enabled="false" />

    <clustering mode="replication" clusterName="CacheCluster">
        
        <async useReplQueue="true" replQueueInterval="10000"
            replQueueMaxElements="500" serializationExecutorPoolSize="20"
            serializationExecutorQueueSize="5000000"/>
        
        <buddy enabled="true" poolName="GoProBuddyPoolReplicationGroup"
            communicationTimeout="20000">
            <dataGravitation auto="true" removeOnFind="true"
                searchBackupTrees="true" />
            <locator>
                <properties>
                    numBuddies=1
                    ignoreColocatedBuddies=true
                </properties>
            </locator>
        </buddy>
        
          <jgroupsConfig>
             <UDP discard_incompatible_packets="true" enable_bundling="false" enable_diagnostics="true" ip_ttl="2"
                  loopback="false" max_bundle_size="64000" max_bundle_timeout="30" mcast_addr="228.10.10.10"
                  mcast_port="45588" mcast_recv_buf_size="25000000" mcast_send_buf_size="640000"
                  oob_thread_pool.enabled="true" oob_thread_pool.keep_alive_time="10000" oob_thread_pool.max_threads="4"
                  oob_thread_pool.min_threads="1" oob_thread_pool.queue_enabled="true" oob_thread_pool.queue_max_size="10"
                  oob_thread_pool.rejection_policy="Run" thread_naming_pattern="pl" thread_pool.enabled="true"
                  thread_pool.keep_alive_time="30000" thread_pool.max_threads="25" thread_pool.min_threads="1"
                  thread_pool.queue_enabled="true" thread_pool.queue_max_size="10" thread_pool.rejection_policy="Run"
                  tos="8" ucast_recv_buf_size="20000000" ucast_send_buf_size="640000" use_concurrent_stack="true"
                  use_incoming_packet_handler="true"/>
             <PING num_initial_members="3" timeout="2000"/>
             <MERGE2 max_interval="30000" min_interval="10000"/>
             <FD_SOCK/>
             <FD max_tries="5" shun="true" timeout="10000"/>
             <VERIFY_SUSPECT timeout="1500"/>
             <pbcast.NAKACK discard_delivered_msgs="true" gc_lag="0" retransmit_timeout="300,600,1200,2400,4800"
                            use_mcast_xmit="false"/>
             <UNICAST timeout="300,600,1200,2400,3600"/>
             <pbcast.STABLE desired_avg_gossip="50000" max_bytes="400000" stability_delay="1000"/>
             <pbcast.GMS join_timeout="5000" print_local_addr="true" shun="false" view_ack_collection_timeout="5000"
                         view_bundling="true"/>
             <FRAG2 frag_size="60000"/>
             <pbcast.STREAMING_STATE_TRANSFER/>
             <pbcast.FLUSH timeout="0"/>

          </jgroupsConfig>

    </clustering>

    <eviction wakeUpInterval="3600">
        <default algorithmClass="org.jboss.cache.eviction.LRUAlgorithm"
            eventQueueSize="200000">
            <property name="maxNodes" value="-1" />
            <property name="timeToLive" value="-1" />
        </default>

        <region name="/DEFAULT_NODE">
        </region>
    </eviction>

</jbosscache>

[svkap]

Hi again.

It seems that the error was in my code. Objects are removed just fine. I have problems replacing objects on the clustered env but I think that it will be better to start a new thread.