RabbitMQ fails to launch during deployment. CentOS

[Kirill Omelchenko]

http://jenkins-product.srt.mirantis.net:8080/view/0_master_swarm/job/master_fuelmain.system_test.centos.thread_4/146/testReport/%28root%29/deploy_ha_flat/deploy_ha_flat/

Configuration:
- HA, Nova FlatDHCP
- 3x Controllers
- 2x Computes

deployment failed due to next errors in OpenStack logs:

2014-08-27 14:42:36 ERROR
oslo.messaging._drivers.impl_rabbit [-] AMQP server on 10.108.32.7:5673 is unreachable: timed out. Trying again in 30 seconds.
2014-08-27 14:42:00 ERROR
oslo.messaging._drivers.impl_rabbit [-] AMQP server on 127.0.0.1:5673 is unreachable: [Errno 111] ECONNREFUSED. Trying again in 30 seconds.
2014-08-27 14:41:29 ERROR
oslo.messaging._drivers.impl_rabbit [-] AMQP server on 10.108.32.6:5673 is unreachable: [Errno 113] EHOSTUNREACH. Trying again in 30 seconds.
2014-08-27 14:40:57 ERROR
oslo.messaging._drivers.impl_rabbit [-] AMQP server on 10.108.32.7:5673 is unreachable: [Errno 113] EHOSTUNREACH. Trying again in 30 seconds.
2014-08-27 14:40:25 ERROR
oslo.messaging._drivers.impl_rabbit [-] AMQP server on 127.0.0.1:5673 is unreachable: [Errno 111] ECONNREFUSED. Trying again in 30 seconds.

[Vladimir Kuklin (vkuklin)]

which service logs is this?

[Kirill Omelchenko (komelchenko)]

It was nova-scheduler for sure. And other schedulers too have similar messages in their logs.

[Vladimir Kuklin (vkuklin)]

troubleshooting shows the following:

http://paste.openstack.org/show/101042/

there is no connectivity between broken controllers and other nodes. Marking the bug as invalid.