Do you guys think switching to Kaha Data Store could solve this issue ( <=> this issue is AmqDatastore specific?)
On 16 March 2010 11:00, Sebastian Rodriguez <[email protected]> wrote: > Dear all, > I come back to you to look for pointers on possible causes that can lead me > to receive this kind of exception. > I'm using Apache.ActiveMQ 5.3.1, with Apache.NMS.ActiveMQ 1.2.0. > I'm using ActiveMQ for communicating in a SOAP application: the servers are > in Network of Brokers configuration, the clients have an embedded ActiveMQ. > > Some of the members of the network of brokers are not always available, and > when reconnecting I get a lot of these kind of exceptions one after the > other: > > INFO | jvm 1 | 2010/03/15 19:12:41 | WARN AMQMessageStore > - Message could not be added to long term store: Index: 1, Size: 1 > INFO | jvm 1 | 2010/03/15 19:12:41 | > java.lang.IndexOutOfBoundsException: Index: 1, Size: 1 > INFO | jvm 1 | 2010/03/15 19:12:41 | at > java.util.ArrayList.RangeCheck(Unknown Source) > INFO | jvm 1 | 2010/03/15 19:12:41 | at > java.util.ArrayList.get(Unknown Source) > INFO | jvm 1 | 2010/03/15 19:12:41 | at > org.apache.activemq.kaha.impl.index.hash.HashPage.getHashEntry(HashPage.java:181) > INFO | jvm 1 | 2010/03/15 19:12:41 | at > org.apache.activemq.kaha.impl.index.hash.HashPageInfo.getHashEntry(HashPageInfo.java:78) > INFO | jvm 1 | 2010/03/15 19:12:41 | at > org.apache.activemq.kaha.impl.index.hash.HashBin.getHashEntry(HashBin.java:247) > INFO | jvm 1 | 2010/03/15 19:12:41 | at > org.apache.activemq.kaha.impl.index.hash.HashBin.remove(HashBin.java:164) > INFO | jvm 1 | 2010/03/15 19:12:41 | at > org.apache.activemq.kaha.impl.index.hash.HashIndex.remove(HashIndex.java:303) > INFO | jvm 1 | 2010/03/15 19:12:41 | at > org.apache.activemq.kaha.impl.container.MapContainerImpl.remove(MapContainerImpl.java:337) > INFO | jvm 1 | 2010/03/15 19:12:41 | at > org.apache.activemq.kaha.impl.container.MapContainerImpl.place(MapContainerImpl.java:414) > INFO | jvm 1 | 2010/03/15 19:12:41 | at > org.apache.activemq.store.kahadaptor.TopicSubContainer.add(TopicSubContainer.java:62) > INFO | jvm 1 | 2010/03/15 19:12:41 | at > org.apache.activemq.store.kahadaptor.KahaTopicReferenceStore.addMessageReference(KahaTopicReferenceStore.java:104) > INFO | jvm 1 | 2010/03/15 19:12:41 | at > org.apache.activemq.store.amq.AMQMessageStore$4.execute(AMQMessageStore.java:383) > INFO | jvm 1 | 2010/03/15 19:12:41 | at > org.apache.activemq.util.TransactionTemplate.run(TransactionTemplate.java:44) > INFO | jvm 1 | 2010/03/15 19:12:41 | at > org.apache.activemq.store.amq.AMQMessageStore.doAsyncWrite(AMQMessageStore.java:373) > INFO | jvm 1 | 2010/03/15 19:12:41 | at > org.apache.activemq.store.amq.AMQMessageStore.asyncWrite(AMQMessageStore.java:339) > INFO | jvm 1 | 2010/03/15 19:12:41 | at > org.apache.activemq.store.amq.AMQMessageStore$1.iterate(AMQMessageStore.java:90) > INFO | jvm 1 | 2010/03/15 19:12:41 | at > org.apache.activemq.thread.DedicatedTaskRunner.runTask(DedicatedTaskRunner.java:98) > INFO | jvm 1 | 2010/03/15 19:12:41 | at > org.apache.activemq.thread.DedicatedTaskRunner$1.run(DedicatedTaskRunner.java:36) > > Until at some point I get this message: > > 2010-03-16 09:19:44,608 [216.10.21:31616] WARN DemandForwardingBridge > - Network connection between vm://client1 and tcp://broker1:31616 > shutdown due to a remote error: java.io.EOFException > 2010-03-16 09:19:44,623 [216.10.21:31616] INFO DemandForwardingBridge > - client1 bridge to client2 stopped > 2010-03-16 09:19:44,623 [r$wor...@c0bc3a] INFO DiscoveryNetworkConnector > - Establishing network connection from vm://client1 to > tcp://broker1:31616 > 2010-03-16 09:19:45,826 [r$wor...@c0bc3a] WARN DiscoveryNetworkConnector > - Could not start network bridge between: vm://client1 and: > tcp://broker1:31616 due to: java.net.ConnectException: Connection refused: > connect > > Then it tries to reconnect and reconnect > At that point the Service is "hang", meaning the service seems to be > running, yet when looking for connections, they are all gone. My client will > not receive any more messages from the server. > > After rebooting the service, it will take incredibly long to recover. This > is caused I guess also because right now I'm using the AMQDataStore (I > believe now it's recommended to switch to the KahaStore). > > What can cause this message problem and is it the reason that after a lot > of these exceptions i receive the EOF Exception? > > Thank you very much for your help, > -- > Sebastien Rodriguez > -- Sebastien Rodriguez
