Description
Our kafka cluster encountered a few disk/xfs failures in the cloud vm instances. When a disk/xfs failure happens, kafka process did not exit gracefully. Instead, it ran into "<defunct>" status, with port 9092 still be reachable. when failures like this happens, kafka should shutdown all threads and exit. The following is the kafka logs when the failure happens:
[2018-02-08 12:52:31,764] ERROR Error while accepting connection (kafka.network.Acceptor) java.io.IOException: Too many open files at sun.nio.ch.ServerSocketChannelImpl.accept0(Native Method) at sun.nio.ch.ServerSocketChannelImpl.accept(ServerSocketChannelImpl.java:422) at sun.nio.ch.ServerSocketChannelImpl.accept(ServerSocketChannelImpl.java:250) at kafka.network.Acceptor.accept(SocketServer.scala:340) at kafka.network.Acceptor.run(SocketServer.scala:283) at java.lang.Thread.run(Thread.java:748) [2018-02-08 12:52:31,772] ERROR Error while accepting connection (kafka.network.Acceptor) java.io.IOException: Too many open files at sun.nio.ch.ServerSocketChannelImpl.accept0(Native Method) at sun.nio.ch.ServerSocketChannelImpl.accept(ServerSocketChannelImpl.java:422) at sun.nio.ch.ServerSocketChannelImpl.accept(ServerSocketChannelImpl.java:250) at kafka.network.Acceptor.accept(SocketServer.scala:340) at kafka.network.Acceptor.run(SocketServer.scala:283) at java.lang.Thread.run(Thread.java:748) [2018-02-08 12:52:31,772] ERROR Error while accepting connection (kafka.network.Acceptor) java.io.IOException: Too many open files at sun.nio.ch.ServerSocketChannelImpl.accept0(Native Method) at sun.nio.ch.ServerSocketChannelImpl.accept(ServerSocketChannelImpl.java:422) at sun.nio.ch.ServerSocketChannelImpl.accept(ServerSocketChannelImpl.java:250) at kafka.network.Acceptor.accept(SocketServer.scala:340) at kafka.network.Acceptor.run(SocketServer.scala:283) at java.lang.Thread.run(Thread.java:748)