AnsweredAssumed Answered

restore cluster after failed cldb server

Question asked by gabi_k on Oct 2, 2011
Latest reply on Oct 6, 2011 by srivas
Hi,

after my cldb failed server wend down, I installed zookeeper on another node (of 3) and configure it to be cldb and zookeeper (with configure.sh), after that I instaled cldb on this node (so I have 1 down and 2 up, 1 of then now have cldb and zk).

I am trying to start the cluster but I got the following error:
cldb.log:
....

2011-10-02 23:00:36,903 INFO  com.mapr.fs.license.LicenseManager [pool-1-thread-2]: using default validator 
2011-10-02 23:00:36,909 INFO  com.mapr.kvstore.Operation [pool-1-thread-2]: setNoDelete=true 
2011-10-02 23:00:37,171 INFO  com.mapr.fs.license.LicenseManager [pool-1-thread-2]: 0x7: unique id: 000c297ee45b-564D927B-7A2A-D99E-EEF5-65889D7EE45B-0000913921-0000652802-0000261122 
2011-10-02 23:00:37,171 INFO  com.mapr.fs.cldb.CLDBServer [pool-1-thread-2]: CLDB HA check returned: true 
2011-10-02 23:00:37,171 INFO  com.mapr.fs.cldb.CLDBServer [pool-1-thread-2]: CLDB Mode MASTER_REGISTER_READY have master now 
2011-10-02 23:00:37,171 INFO  com.mapr.fs.cldb.CLDBServer [pool-1-thread-2]: CLDB Mode MASTER_REGISTER_READY Init Table 
2011-10-02 23:00:37,172 INFO  com.mapr.kvstore.Operation [pool-1-thread-2]: setNoDelete=true
Got fidmsg 
2011-10-02 23:00:37,212 INFO  com.mapr.fs.cldb.CLDBServer [pool-1-thread-2]: CLDB can become a read write copy now 
2011-10-02 23:00:37,212 INFO  com.mapr.fs.cldb.table.Table [pool-1-thread-2]: KvStore Init : Creating Tables 
2011-10-02 23:00:37,255 FATAL com.mapr.fs.cldb.CLDBServer [pool-1-thread-2]: CLDB running with stale ZooKeeper information.  ZkEpoch: 3 TableEpoch: 18 Please restore ZooKeeper state and restart CLDB 
2011-10-02 23:00:37,255 FATAL com.mapr.fs.cldb.CLDB [pool-1-thread-2]: CLDBShutdown: Shutting down CLDB because of stale ZooKeeper information. 
2011-10-02 23:00:37,255 INFO  com.mapr.fs.cldb.CLDBServer [pool-1-thread-2]: Shutdown: Stopping CLDB 
2011-10-02 23:00:37,256 INFO  com.mapr.fs.cldb.CLDB [Thread-9]: CLDB shutdown 


What can I do know to start it?

Thanks,

Outcomes