Kafka broker unable to recover - apache-kafka

Question

I am running a 5 node kafka broker with a 5 node zookeeper ensemble.
On Kafka nodes I keep getting this error
[2016-11-14 19:05:11,345] INFO Opening socket connection to server 10.105.23.188/10.105.23.188:2181. Will not attempt to authenticate using SASL (unknown error) (org.apache.zookeeper.ClientCnxn)
[2016-11-14 19:05:11,345] INFO Socket connection established to 10.105.23.188/10.105.23.188:2181, initiating session (org.apache.zookeeper.ClientCnxn)
[2016-11-14 19:05:11,346] INFO Unable to read additional data from server sessionid 0x55861d8ea6f0000, likely server has closed socket, closing socket connection and attempting reconnect (org.apache.zookeeper.ClientCnxn)
[2016-11-14 19:05:11,721] INFO Opening socket connection to server 10.105.25.4/10.105.25.4:2181. Will not attempt to authenticate using SASL (unknown error) (org.apache.zookeeper.ClientCnxn)
[2016-11-14 19:05:11,721] INFO Socket connection established to 10.105.25.4/10.105.25.4:2181, initiating session (org.apache.zookeeper.ClientCnxn)
[2016-11-14 19:05:11,722] INFO Unable to read additional data from server sessionid 0x55861d8ea6f0000, likely server has closed socket, closing socket connection and attempting reconnect (org.apache.zookeeper.ClientCnxn)
[2016-11-14 19:05:11,841] INFO Opening socket connection to server 10.105.24.4/10.105.24.4:2181. Will not attempt to authenticate using SASL (unknown error) (org.apache.zookeeper.ClientCnxn)
[2016-11-14 19:05:11,841] WARN Session 0x55861d8ea6f0000 for server null, unexpected error, closing socket connection and attempting reconnect (org.apache.zookeeper.ClientCnxn)
java.net.ConnectException: Connection refused
at sun.nio.ch.SocketChannelImpl.checkConnect(Native Method)
at sun.nio.ch.SocketChannelImpl.finishConnect(SocketChannelImpl.java:717)
at org.apache.zookeeper.ClientCnxnSocketNIO.doTransport(ClientCnxnSocketNIO.java:361)
at org.apache.zookeeper.ClientCnxn$SendThread.run(ClientCnxn.java:1081)
[2016-11-14 19:05:12,319] INFO Opening socket connection to server 10.105.27.23/10.105.27.23:2181. Will not attempt to authenticate using SASL (unknown error) (org.apache.zookeeper.ClientCnxn)
[2016-11-14 19:05:12,319] INFO Socket connection established to 10.105.27.23/10.105.27.23:2181, initiating session (org.apache.zookeeper.ClientCnxn)
[2016-11-14 19:05:12,320] INFO Unable to read additional data from server sessionid 0x55861d8ea6f0000, likely server has closed socket, closing socket connection and attempting reconnect (org.apache.zookeeper.ClientCnxn)
[2016-11-14 19:05:12,395] WARN [ReplicaFetcherThread-0-5], Error in fetch kafka.server.ReplicaFetcherThread$FetchRequest#3fbe518f. Possible cause: org.apache.kafka.common.protocol.types.SchemaException: Error reading field 'responses': Error reading array of size 1204063, only 920 bytes available (kafka.server.ReplicaFetcherThread)
On Zookeeper nodes :
I get the following on the leader
2016-11-14 19:14:47,502 [myid:4] - ERROR [LearnerHandler-/10.105.25.4:50099:LearnerHandler#631] - Unexpected exception causing shutdown while sock still open
java.io.EOFException
at java.io.DataInputStream.readInt(DataInputStream.java:392)
at org.apache.jute.BinaryInputArchive.readInt(BinaryInputArchive.java:63)
at org.apache.zookeeper.server.quorum.QuorumPacket.deserialize(QuorumPacket.java:83)
at org.apache.jute.BinaryInputArchive.readRecord(BinaryInputArchive.java:99)
at org.apache.zookeeper.server.quorum.LearnerHandler.run(LearnerHandler.java:513)
When a consumer tries to read from this cluster it get this error repeatedly
2016-11-14 19:17:36,847 WARN {main} [kafka.clients.NetworkClient$DefaultMetadataUpdater:handleResponse:629] Error while fetching metadata with correlation id 1603 : {=UNKNOWN_TOPIC_OR_PARTITION}
Please help me debug this. The consumers have auto commit enabled. If you need me to provide more configuration please let me know.
Thanks!