报错：Unable to read additional data from client sessionid 0x36ab52d38c20b20, likely client has closed socket

报错背景：

CDH集群中，将kafka和Flume整合，将kafka的数据发送给Flume消费。

启动kafka的时候正常，但是启动Flume的时候出现了报错现象。

但是我检查了Flume，Flume的状态很正常，于是我又查看了zookeeper的日志（/var/log/zookeeper/），发现了报错。

报错现象：

-- ::, INFO org.apache.zookeeper.server.NIOServerCnxnFactory: Accepted socket connection from /192.168.52.26:
-- ::, INFO org.apache.zookeeper.server.ZooKeeperServer: Client attempting to establish new session at /192.168.52.26:
-- ::, INFO org.apache.zookeeper.server.ZooKeeperServer: Established session 0x36ab52d38c20b20 with negotiated timeout  for client /192.168.52.26:
-- ::, WARN org.apache.zookeeper.server.NIOServerCnxn: caught end of stream exception
EndOfStreamException: Unable to read additional data from client sessionid 0x36ab52d38c20b20, likely client has closed socket
    at org.apache.zookeeper.server.NIOServerCnxn.doIO(NIOServerCnxn.java:)
    at org.apache.zookeeper.server.NIOServerCnxnFactory.run(NIOServerCnxnFactory.java:)
    at java.lang.Thread.run(Thread.java:)
-- ::, INFO org.apache.zookeeper.server.NIOServerCnxn: Closed socket connection for client /192.168.52.26: which had sessionid 0x36ab52d38c20b20
-- ::, INFO org.apache.zookeeper.server.NIOServerCnxnFactory: Accepted socket connection from /192.168.52.26:

报错原因：

原因不明。

但是这个报错肯定和kafka报错有关系。

网上的报错解决办法都是加长zookeeper对kafka的心跳检测时间，但是这并不能解决问题。可能是因为kafka的broker已经死掉了，所以心跳检测会失败。

报错解决：

这个报错其实不算是一个严重报错，它是由zookeeper管理的其它组件报错引起的，只要把对应报错组件的报错解决，此报错就会消失。（以上内容全为个人推测）