Comment 16 for bug 1373569

@Tyler you can restart them by 'crm resource restart p_foo_name' where the p_foo_name is a name of corresponding resource in 'crm status'

Anyway, looks like the 2nd discovered issue has nothing to the sockets but with amqp channels, i.e. rabbitmq cluster health.
If I issue 'rabbitmqctl list_channels', it hangs for ever - you can check it as well.

Normally, if you restart the rabbitmq cluster with 'crm resource restart master_p_rabbitmq-server' and give him ~5 min to reassemble completely, the problem with hanging list_channels should gone. Note, that that action would cause *complete* downtime for your cloud unless rabbit cluster reassembled. You can check whether rabbit cluster reassembled by 'fuel health --env 3 --check ha'