hbase运维笔记

（hbase0.98.7 + hadoop2.5.0）

hbase启动后，master马上挂掉，错误很诡异：

2014-12-10 15:45:59,239 DEBUG [master:XX-249-83:60000] zookeeper.ZKAssign: master:60000-0x34a32d840170020, quorum=XX-249-75:2181,XX-249-74:2181,XX-249-76:2181, baseZNode=/hbase Creating (or updating) unassigned node 1588230740 with OFFLINE state
2014-12-10 15:45:59,259 DEBUG [master:XX-249-83:60000] master.AssignmentManager: Setting table hbase:meta to ENABLED state.
2014-12-10 15:45:59,274 INFO  [master:XX-249-83:60000] master.AssignmentManager: Assigning hbase:meta,,1.1588230740 to XX-249-77,60020,1418197540838
2014-12-10 15:45:59,274 INFO  [master:XX-249-83:60000] master.RegionStates: Transition {1588230740 state=OFFLINE, ts=1418197559238, server=null} to {1588230740 state=PENDING_OPEN, ts=1418197559274, server=XX-249-77,60020,1418197540838}?
2014-12-10 15:45:59,274 DEBUG [master:XX-249-83:60000] master.ServerManager: New admin connection to XX-249-77,60020,1418197540838
2014-12-10 15:45:59,794 INFO  [master:XX-249-83:60000] master.ServerManager: AssignmentManager hasn't finished failover cleanup; waiting
2014-12-10 15:45:59,794 FATAL [master:XX-249-83:60000] master.HMaster: Master server abort: loaded coprocessors are: []
2014-12-10 15:45:59,796 FATAL [master:XX-249-83:60000] master.HMaster: Unhandled exception. Starting shutdown.
java.lang.IllegalAccessError: tried to access method com.google.common.base.Stopwatch.()V from class org.apache.hadoop.hbase.catalog.CatalogTracker
        at org.apache.hadoop.hbase.catalog.CatalogTracker.waitForMeta(CatalogTracker.java:307)
        at org.apache.hadoop.hbase.master.HMaster.enableServerShutdownHandler(HMaster.java:1074)
        at org.apache.hadoop.hbase.master.HMaster.assignMeta(HMaster.java:1036)
        at org.apache.hadoop.hbase.master.HMaster.finishInitialization(HMaster.java:888)
        at org.apache.hadoop.hbase.master.HMaster.run(HMaster.java:605)
        at java.lang.Thread.run(Thread.java:745)
2014-12-10 15:45:59,799 INFO  [master:XX-249-83:60000] master.HMaster: Aborting
2014-12-10 15:45:59,799 DEBUG [master:XX-249-83:60000] master.HMaster: Stopping service threads
2014-12-10 15:45:59,799 INFO  [master:XX-249-83:60000] ipc.RpcServer: Stopping server on 60000
2014-12-10 15:45:59,799 INFO  [RpcServer.listener,port=60000] ipc.RpcServer: RpcServer.listener,port=60000: stopping
2014-12-10 15:45:59,800 INFO  [RpcServer.responder] ipc.RpcServer: RpcServer.responder: stopped
2014-12-10 15:45:59,800 INFO  [RpcServer.responder] ipc.RpcServer: RpcServer.responder: stopping

然后查看了源码，试了很多办法都不行，最后无意中发现hadoop中的guava包跟hbase中的不一致，把hbase中的替换掉就ok了。