-
1. Re: Buddy location issue
thogha Oct 25, 2009 8:05 PM (in response to thogha)Sorry, I wiped out my original post in the preview.
I'm trying to get buddy replication working. I'm running two servers, each with one instance of 3.1.0 inside Tomcat. Here is the error (from catalina.out):Oct 25, 2009 7:26:45 PM org.jgroups.JChannel init INFO: JGroups version: 2.6.7.GA ------------------------------------------------------- GMS: address is 127.0.1.1:7800 ------------------------------------------------------- Oct 25, 2009 7:26:48 PM org.jboss.cache.RPCManagerImpl$MembershipListenerAdaptor viewAccepted INFO: Received new cluster view: [127.0.1.1:7800|0] [127.0.1.1:7800] Oct 25, 2009 7:26:48 PM org.jboss.cache.RPCManagerImpl start INFO: Cache local address is 127.0.1.1:7800 Oct 25, 2009 7:26:48 PM org.jboss.cache.buddyreplication.NextMemberBuddyLocator locateBuddies INFO: Expected to look for 1 buddies but could only find 0 suitable candidates - trying with colocated buddies as well. Oct 25, 2009 7:26:48 PM org.jboss.cache.buddyreplication.NextMemberBuddyLocator locateBuddies INFO: Expected to look for 1 buddies but could only find 0 suitable candidates - trying again, ignoring buddy pool hints. Oct 25, 2009 7:26:48 PM org.jboss.cache.buddyreplication.NextMemberBuddyLocator locateBuddies INFO: Expected to look for 1 buddies but could only find 0 suitable candidates - trying with colocated buddies as well. Oct 25, 2009 7:26:48 PM org.jboss.cache.buddyreplication.NextMemberBuddyLocator locateBuddies INFO: Expected to look for 1 buddies but could only find 0 suitable candidates!
And here is JBossCacheConfiguration.xml<?xml version="1.0" encoding="UTF-8"?> <jbosscache xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xmlns="urn:jboss:jbosscache-core:config:3.1"> <eviction wakeUpInterval="500"> <default algorithmClass="org.jboss.cache.eviction.LFUAlgorithm" eventQueueSize="1000000"> <property name="minNodes" value="1000000" /> <property name="maxNodes" value="2000000" /> </default> </eviction> <clustering mode="invalidation" clusterName="cluster1"> <sync /> <buddy enabled="true" poolName="buddyPool1" communicationTimeout="2000"> <dataGravitation auto="false" removeOnFind="false" searchBackupTrees="false" /> <locator class="org.jboss.cache.buddyreplication.NextMemberBuddyLocator"> <properties> numBuddies = 1 ignoreColocatedBuddies = true </properties> </locator> </buddy> <jgroupsConfig> <TCP start_port="7800" discard_incompatible_packets="true" enable_bundling="false" enable_diagnostics="true" enable_unicast_bundling="false" loopback="false" max_bundle_size="64000" max_bundle_timeout="30" oob_thread_pool.enabled="true" oob_thread_pool.keep_alive_time="10000" oob_thread_pool.max_threads="8" oob_thread_pool.min_threads="2" oob_thread_pool.queue_enabled="false" oob_thread_pool.queue_max_size="10" oob_thread_pool.rejection_policy="Run" recv_buf_size="20000000" thread_naming_pattern="pl" thread_pool.enabled="true" thread_pool.keep_alive_time="30000" thread_pool.max_threads="4" thread_pool.min_threads="1" thread_pool.queue_enabled="true" thread_pool.queue_max_size="10000" thread_pool.rejection_policy="discard" use_concurrent_stack="true" use_send_queues="false" /> <MPING mcast_addr="232.1.2.3" 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" /> <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" /> <FC max_credits="1000000" min_threshold="0.20" /> <FRAG2 frag_size="60000" /> <pbcast.STREAMING_STATE_TRANSFER /> <pbcast.FLUSH timeout="0" /> </jgroupsConfig> </clustering> </jbosscache>
How does one diagnose and solve this problem? -
2. Re: Buddy location issue
manik Oct 29, 2009 9:22 AM (in response to thogha)Looks like there is no one else in your cluster?
-
3. Re: Buddy location issue
thogha Nov 3, 2009 2:23 PM (in response to thogha)Hi! I'm running two servers. Same config each, and same log output from both.
The servers are Xen-provisioned VSP in the same data center. I'm thinking this may be an issue with Xen multicast configuration.
Can JGroups can be configured to use TCP and a list of IP addresses instead of multicast?
Is replication possible in a VPS environment?? -
4. Re: Buddy location issue
belaban Nov 3, 2009 4:14 PM (in response to thogha)127.0.0.1 won't work unless both of your processes run on the same box. Use -Djgroups.bind_addr=192.168.x.x to pick a correct NIC.
Yes, JGroups can use TCP and TCPPING for a static list instead of multicasting, read the details in the manual at jgroups.org.