[rabbitmq-discuss] Inconsistent results on nodes for rabbitmqctl list_queues

[Drew Smathers]

I'm using rabbitmq (1.5.3) in a cluster of nodes and noticed that
rabbitmqctl will sometimes report inconsistent results for list_queues
from one to another.  "status" indicates the cluster is operating
normally, showing all nodes running:

  [{running_applications,[{rabbit,"RabbitMQ","1.5.3"},
                        {mnesia,"MNESIA  CXC 138 12","4.4.3"},
                        {os_mon,"CPO  CXC 138 46","2.1.6"},
                        {sasl,"SASL  CXC 138 11","2.1.5.3"},
                        {stdlib,"ERTS  CXC 138 10","1.15.3"},
                        {kernel,"ERTS  CXC 138 10","2.12.3"}]},
 {nodes,[rabbit at hostA,rabbit at hostB,
         rabbit at hostC,rabbit at hostD]},
 {running_nodes,[rabbit at hostA,rabbit at hostB,
                 rabbit at hostC,rabbit at hostD]}]

Where hostA/B are disc nodes and hostC/D are ram nodes.  However,
running list_queues on hostC shows incosistent results from hostA,B,D.

list_queues on hostA:

  foo1  rabbit at hostA  0
  foo2  rabbit at hostA  0
  foo3  rabbit at hostA  0
  foo4  rabbit at hostA  0

list_queues on hostC:

  foo1  rabbit at hostA  0
  foo3  rabbit at hostA  0
  foo4  rabbit at hostA  0

What kind of scenarios could cause this to occur?  This is of concern
to, me because upon testing declaration of foo2 from hostC,
unexpectedly the queue "migrates" from hostA to hostC (from the
perspective of all hosts) with consumers reset to 0; hence, consumers
on foo2 stop receiving messages.

-Drew