Cloudera HA HDFS NameNode无法启动

Question

当我尝试为Cloudera HDFS服务启用高可用性时，我的NameNode主管无法启动此错误：

java.io.IOException: Cannot start an HA namenode with name dirs that need recovery. Dir: Storage Directory /disk0/dfs/nn state: NOT_FORMATTED

这是在我已经运行之后：hadoop namenode -format

hadoop namenode -format的输出：

STARTUP_MSG:   build = file:///var/lib/jenkins/workspace/generic-package-ubuntu64-12-04/CDH4.2.1-Packaging-Hadoop-2013-04-22_09-50-19/hadoop-2.0.0+960-1.cdh4.2.1.p0.9~precise/src/hadoop-common-project/hadoop-common -r 144bd548d481c2774fab2bec2ac2645d190f705b; compiled by 'jenkins' on Mon Apr 22 10:26:30 PDT 2013
STARTUP_MSG:   java = 1.6.0_38
************************************************************/
13/09/09 21:39:18 WARN common.Util: Path /var/lib/hadoop-hdfs/cache/hdfs/dfs/name should be specified as a URI in configuration files. Please update hdfs configuration.
13/09/09 21:39:18 WARN common.Util: Path /var/lib/hadoop-hdfs/cache/hdfs/dfs/name should be specified as a URI in configuration files. Please update hdfs configuration.
Formatting using clusterid: CID-4d433d8e-2528-4dd7-b6d0-ecdf71096c34
13/09/09 21:39:18 INFO util.HostsFileReader: Refreshing hosts (include/exclude) list
13/09/09 21:39:18 INFO blockmanagement.DatanodeManager: dfs.block.invalidate.limit=1000
13/09/09 21:39:18 INFO blockmanagement.BlockManager: dfs.block.access.token.enable=false
13/09/09 21:39:18 INFO blockmanagement.BlockManager: defaultReplication         = 3
13/09/09 21:39:18 INFO blockmanagement.BlockManager: maxReplication             = 512
13/09/09 21:39:18 INFO blockmanagement.BlockManager: minReplication             = 1
13/09/09 21:39:18 INFO blockmanagement.BlockManager: maxReplicationStreams      = 2
13/09/09 21:39:18 INFO blockmanagement.BlockManager: shouldCheckForEnoughRacks  = false
13/09/09 21:39:18 INFO blockmanagement.BlockManager: replicationRecheckInterval = 3000
13/09/09 21:39:18 INFO blockmanagement.BlockManager: encryptDataTransfer        = false
13/09/09 21:39:18 INFO namenode.FSNamesystem: fsOwner             = root (auth:SIMPLE)
13/09/09 21:39:18 INFO namenode.FSNamesystem: supergroup          = supergroup
13/09/09 21:39:18 INFO namenode.FSNamesystem: isPermissionEnabled = true
13/09/09 21:39:18 INFO namenode.FSNamesystem: HA Enabled: false
13/09/09 21:39:18 INFO namenode.FSNamesystem: Append Enabled: true
13/09/09 21:39:19 INFO namenode.NameNode: Caching file names occuring more than 10 times
13/09/09 21:39:19 INFO namenode.FSNamesystem: dfs.namenode.safemode.threshold-pct = 0.9990000128746033
13/09/09 21:39:19 INFO namenode.FSNamesystem: dfs.namenode.safemode.min.datanodes = 0
13/09/09 21:39:19 INFO namenode.FSNamesystem: dfs.namenode.safemode.extension     = 30000
13/09/09 21:39:20 INFO namenode.NNStorage: Storage directory /var/lib/hadoop-hdfs/cache/hdfs/dfs/name has been successfully formatted.
13/09/09 21:39:20 INFO namenode.FSImage: Saving image file /var/lib/hadoop-hdfs/cache/hdfs/dfs/name/current/fsimage.ckpt_0000000000000000000 using no compression
13/09/09 21:39:20 INFO namenode.FSImage: Image file of size 119 saved in 0 seconds.
13/09/09 21:39:20 INFO namenode.NNStorageRetentionManager: Going to retain 1 images with txid >= 0
13/09/09 21:39:20 INFO util.ExitUtil: Exiting with status 0
13/09/09 21:39:20 INFO namenode.NameNode: SHUTDOWN_MSG:
/************************************************************
SHUTDOWN_MSG: Shutting down NameNode at *IP_ADDRESS*
************************************************************/

有没有人有任何想法？

Answer 1

如果您一直在启动备用名称节点，您可以尝试：

hadoop namenod -bootstrapStandby

...在启动namenode之前。