mxosrvr core when zookeeper connection gets dropped due to timeout

[Aruna Sadashiva]

Mxosrvr cores were seen on Zircon during perf tests. the connection to zookeeper is getting dropped (zh=0x0) – maybe because of timeout or other errors.

#0 0x00007ffff4a318a5 in raise () from /lib64/libc.so.6
#1 0x00007ffff4a3300d in abort () from /lib64/libc.so.6
#2 0x00007ffff5d50a55 in os::abort(bool) ()
   from /usr/java/jdk1.7.0_67/jre/lib/amd64/server/libjvm.so
#3 0x00007ffff5ed0f87 in VMError::report_and_die() ()
   from /usr/java/jdk1.7.0_67/jre/lib/amd64/server/libjvm.so
#4 0x00007ffff5d5596f in JVM_handle_linux_signal ()
   from /usr/java/jdk1.7.0_67/jre/lib/amd64/server/libjvm.so
#5 <signal handler called>
#6 0x00007ffff65bcce9 in zoo_exists (zh=0x0,
    path=0xeb2598 "/squser4/dcs/servers/registered/zircon-n018.usa.hp.com:8:88", watch=0, stat=0x7fffe61bf080) at src/zookeeper.c:3503
#7 0x00000000004c59f5 in updateZKState (currState=CONNECTED,
    newState=AVAILABLE) at SrvrConnect.cpp:9057
#8 0x00000000004ca966 in odbc_SQLSvc_TerminateDialogue_ame_ (
    objtag_=0xedb8b0, call_id_=0xedb908, dialogueId=313965727)
    at SrvrConnect.cpp:3885
#9 0x0000000000493dce in DISPATCH_TCPIPRequest (objtag_=0xedb8b0,
    call_id_=0xedb908, operation_id=<value optimized out>)
    at Interface/odbcs_srvr.cpp:1772
#10 0x0000000000433882 in BUILD_TCPIP_REQUEST (pnode=0xedb8b0)
    at ../Common/TCPIPSystemSrvr.cpp:603
#11 0x000000000043421d in PROCESS_TCPIP_REQUEST (pnode=0xedb8b0)
---Type <return> to continue, or q <return> to quit---
    at ../Common/TCPIPSystemSrvr.cpp:581
#12 0x0000000000462406 in CNSKListenerSrvr::tcpip_listener (arg=0xda5510)
    at Interface/linux/Listener_srvr_ps.cpp:400
#13 0x00007ffff47e52e0 in sb_thread_sthr_disp (pp_arg=0xeb4c70)
    at threadl.cpp:253
#14 0x00007ffff45b1851 in start_thread () from /lib64/libpthread.so.0
#15 0x00007ffff4ae790d in clone () from /lib64/libc.so.6