Elasticsearch Java API 的使用（23）—实现集群管理

集群管理
使用 ClusterHealthResponse 实现集群信息监测和管理
ClusterHealthResponse 的实例 healths 可调用的方法分析：

• 集群健康信息总览：healths.toString()
• 集群名称：healths.getClusterName()
• 集群状态：healths.getStatus()
• 集群是否超时：healths.isTimedOut()
• 集群的节点个数：healths.getNumberOfNodes()
• 集群的数据节点个数：healths.getNumberOfDataNodes()
• 集群的活跃私有分片个数：healths.getActivePrimaryShards()
• 集群的活跃分片个数：healths.getActiveShards()
• 集群的迁移分片个数：healths.getRelocatingShards()
• 集群的初始化分片个数： healths.getInitializingShards()
• 集群的未分配分片个数： healths.getUnassignedShards()
• 集群的延迟未分配分片个数：healths.getDelayedUnassignedShards()
• 集群的等待任务的个数：healths.getNumberOfPendingTasks()
• 集群的快速查询数：healths.getNumberOfInFlightFetch()
• 集群的任务最大队列等待毫秒数：healths.getTaskMaxWaitingTime()
• 集群的活跃分片百分比：healths.getActiveShardsPercent()
• 集群的索引的详细信息：healths.getIndices()
  • 它的结构 Map<Sirng, ClusterIndexHealth> 得键值对；
  • key 是索引，value 是包含索引健康信息的实例

public class EsClusterHealth throws UnknownHostException{
    public void ClusterHealth(TransportClient client){
       ClusterHealthResponse healths = client.admin().cluster().prepareHealth().get();
       //集群名称
       String clusterName = healths.getClusterName();
       System.out.println("clusterName="+clusterName);
	
	   //数据节点的个数
	   int numberOfDataNodes = healths.getNumberOfDateNodes();
	   System.out.println("numberOfDataNodes="+numberOfDataNodes);

	   //节点的个数
	   int numberOfNodes = healths.getNumberOfNodes();
	   System.out.println("numberOfNodes="+numberOfNodes);
		
	   for(ClusterIndexHealth health : healths.getIndices().value()){
		   //循环查看索引
		   String index = health.getIndex();
		   //查看索引的分片数
		   int numberOfShards = health.getNumberOfShards();
		   //查看索引的副本数
		   int numberOfReplicas = health.getNumberOfReplicas();
		   System.out.println("index=%s,numberOfShards=%d,numberOfReplicas=%d\n", index, numberOfShards, numberOfReplicas);
			
		   //查看集群集群状态
		   ClusterHealthStatus status = health.getStatus();
		   System.out.println(status.toString());
	   }
    }
}