Zookeeper errors in local provider cause strange status view and possibly broken topology

I saw these errors in the unit.log, and after that, status transitioned in a most unexpected way:

clint@clint-MacBookPro:~$ juju destroy-service ceph
2011-10-16 11:34:51,764 INFO Service 'ceph' destroyed.
2011-10-16 11:34:51,764 INFO 'destroy_service' command finished successfully
clint@clint-MacBookPro:~$ juju deploy --repository charms local:ceph
2011-10-16 11:35:06,764 INFO Charm deployed as service: 'ceph'
2011-10-16 11:35:06,764 INFO 'deploy' command finished successfully
clint@clint-MacBookPro:~$ juju add-unit ceph
2011-10-16 11:36:27,898 INFO Unit 'ceph/13' added to service 'ceph'
2011-10-16 11:36:27,900 INFO 'add_unit' command finished successfully
clint@clint-MacBookPro:~$ juju add-unit ceph
2011-10-16 11:36:29,621 INFO Unit 'ceph/14' added to service 'ceph'
2011-10-16 11:36:29,623 INFO 'add_unit' command finished successfully
clint@clint-MacBookPro:~$ juju status
machines:
  0: {dns-name: localhost, instance-id: local}
services:
  ceph:
    charm: local:oneiric/ceph-39
    relations: {mds: ceph, mon: ceph, osd: ceph, ssh: ceph}
    units:
      ceph/12:
        machine: 0
        public-address: 192.168.122.72
        relations:
          mds: {state: up}
          mon: {state: up}
          osd: {state: up}
          ssh: {state: up}
        state: started
      ceph/13:
        machine: 0
        public-address: null
        relations: {}
        state: null
      ceph/14:
        machine: 0
        public-address: null
        relations: {}
        state: null
2011-10-16 11:36:35,776 INFO 'status' command finished successfully
clint@clint-MacBookPro:~$ sudo tail -f .juju/data/clint-local/units/ceph-12/unit.log
2011-10-16 18:36:26,134: unit.relation.lifecycle@DEBUG: started relation:mon lifecycle
2011-10-16 18:36:26,142: statemachine@DEBUG: relationworkflowstate: transition complete start (state up) {}
2011-10-16 18:36:26,143: unit.lifecycle@DEBUG: started unit lifecycle
2011-10-16 18:36:26,176: statemachine@DEBUG: unitworkflowstate: transition complete start (state started) {}
2011-10-16 18:36:36,291:469(0x7f5d29462700):ZOO_ERROR@handle_socket_error_msg@1528: Socket [192.168.122.1:56037] zk retcode=-7, errno=110(Connection timed out): connection timed out (exceeded timeout by 5ms)
2011-10-16 18:36:39,784:469(0x7f5d29462700):ZOO_WARN@zookeeper_interest@1461: Exceeded deadline by 3496ms
2011-10-16 18:36:46,453:469(0x7f5d29462700):ZOO_ERROR@handle_socket_error_msg@1528: Socket [192.168.122.1:56037] zk retcode=-7, errno=110(Connection timed out): connection timed out (exceeded timeout by 3ms)
2011-10-16 18:36:49,788:469(0x7f5d29462700):ZOO_WARN@zookeeper_interest@1461: Exceeded deadline by 3338ms
2011-10-16 18:36:49,788:469(0x7f5d29462700):ZOO_INFO@check_events@1585: initiated connection to server [192.168.122.1:56037]
2011-10-16 18:36:49,789:469(0x7f5d29462700):ZOO_ERROR@handle_socket_error_msg@1621: Socket [192.168.122.1:56037] zk retcode=-112, errno=116(Stale NFS file handle): sessionId=0x1330dd27c6e002f has expired.
^Cclint@clint-MacBookPro:~$ juju status
machines:
  0: {dns-name: localhost, instance-id: local}
services:
  ceph:
    charm: local:oneiric/ceph-39
    relations: {mds: ceph, mon: ceph, osd: ceph, ssh: ceph}
    units:
      ceph/12:
        machine: 0
        public-address: 192.168.122.72
        relations: {}
        state: started
      ceph/13:
        machine: 0
        public-address: 192.168.122.94
        relations:
          mds: {state: up}
          mon: {state: up}
          osd: {state: up}
          ssh: {state: up}
        state: started
      ceph/14:
        machine: 0
        public-address: 192.168.122.208
        relations:
          mds: {state: up}
          mon: {state: up}
          osd: {state: up}
          ssh: {state: up}
        state: started
2011-10-16 11:39:31,541 INFO 'status' command finished successfully
clint@clint-MacBookPro:~$
clint@clint-MacBookPro:~$

Will attach the unit.logs from all 3 and the charm as a zip