[HBASE-13314] Fix NPE in HMaster.getClusterStatus()

• Type: Bug
• Status: Closed
• Priority: Minor
• Resolution: Fixed
• Fix Version/s: 1.0.1, 1.1.0, 2.0.0
• Assignee: Matteo Bertozzi
• Reporter: Matteo Bertozzi
• Created: 22/Mar/15 20:31
• Updated: 28/Apr/15 19:19
• Resolved: 24/Mar/15 09:36

Description

On a test cluster we got a

java.lang.NullPointerException
  at org.apache.hadoop.hbase.master.HMaster.getClusterStatus(HMaster.java:1742)
  at org.apache.hadoop.hbase.master.balancer.ClusterStatusChore.chore(ClusterStatusChore.java:50)
  at org.apache.hadoop.hbase.Chore.run(Chore.java:87)

In HMaster.getClusterStatus() we have a couple of NPE.
listChildrenNoWatch() may return null if the node is not found

try {
  backupMasterStrings = ZKUtil.listChildrenNoWatch(this.zooKeeper,
   this.zooKeeper.backupMasterAddressesZNode); <--- THIS MAY RETURN NULL
} catch (KeeperException e) {
  LOG.warn(this.zooKeeper.prefix("Unable to list backup servers"), e);
  backupMasterStrings = new ArrayList<String>(0);
}
List<ServerName> backupMasters = new ArrayList<ServerName>(
      backupMasterStrings.size()); <--- WE DON'T CHECK FOR NULL

then below, we build ClusterStatus with args that may be null

String clusterId = fileSystemManager != null ?
  fileSystemManager.getClusterId().toString() : null;
Map<String, RegionState> regionsInTransition = assignmentManager != null ?
  assignmentManager.getRegionStates().getRegionsInTransition() : null;
String[] coprocessors = cpHost != null ? getMasterCoprocessors() : null;
Map<ServerName, ServerLoad> onlineServers = null;
Set<ServerName> deadServers = null;
if (serverManager != null) {
  deadServers = serverManager.getDeadServers().copyServerNames();
  onlineServers = serverManager.getOnlineServers();
}
return new ClusterStatus(VersionInfo.getVersion(), clusterId,
  onlineServers, deadServers, serverName, backupMasters,
  regionsInTransition, coprocessors, balancerOn);

In ClusterStatus equals(), hashCode(), toString() we don't check for nulls