Radosgw backends is in down state after deployment with errors related to keyring in rados.log

[Tatyanka]

Steps:
1. Deploy ceph cluster with radosgw
2. When cluster ready - run ostf test for haproxy back-ends or run bellow command on controller
haproxy-status.sh | egrep -v "BACKEND|FRONTEND"

radosgw node-5 Status: DOWN/L4CON Sessions: 0 Rate: 0
radosgw node-6 Status: DOWN/L4CON Sessions: 0 Rate: 0
radosgw node-1 Status: DOWN/L4CON Sessions: 0 Rate: 0

radosgw logs looks like
2015-11-04 16:30:52.729338 7f90de2f7700 -1 failed to list objects pool_iterate returned r=-2
2015-11-04 16:30:52.729367 7f90de2f7700 0 ERROR: lists_keys_next(): ret=-2
2015-11-04 16:30:52.729470 7f90de2f7700 0 ERROR: sync_all_users() returned ret=-2

We add ostf tests with expectation that in neutron env all haproxy backens are running, so bvt became red (https://github.com/openstack/fuel-ostf/commit/42311445ad4f2a4a82050d724c53ed0cd7cb94b5)

https://172.18.160.103/job/8.0.ubuntu.bvt_2/113/

[OpenStack Infra (hudson-openstack)]

Fix proposed to branch: master
Review: https://review.openstack.org/241783

[OpenStack Infra (hudson-openstack)]

Reviewed: https://review.openstack.org/241783
Committed: https://git.openstack.org/cgit/openstack/fuel-ostf/commit/?id=b360e411a7cf251294542c4512b601c597a5e957
Submitter: Jenkins
Branch: master

commit b360e411a7cf251294542c4512b601c597a5e957
Author: Tatyana Leontovich <email address hidden>
Date: Thu Nov 5 00:21:16 2015 +0200

    Add radosgw to ignored services in haproxy check

    We have red bvt according to radosgw
    backend in haproxy is in DOWN state

    Bug: 1513237
    Change-Id: I6f5f2fb7e7ea9a799fa2eec620f4cbcd1699ccb7