Reproduced on {"build_id": "2014-04-15_01-37-24", "mirantis": "yes", "build_number": "103", "nailgun_sha": "75ac9d49d6918cf5339f1c1726c8831f3c28f524", "production": "dev", "ostf_sha": "1ac04769cc8034438ccf1176860b25d0f5601c4a", "fuelmain_sha": "a9d39b53bf64b405f056f6acfe08f2055b92391f", "astute_sha": "0cb0b26773bd269000a14748e8ba49419462ecaf", "release": "5.0", "fuellib_sha": "d4299afd31992482888673be4e23936ce91d532a"}
Steps of the test:
1. Create next cluster - CentOS, HA, KVM, Flat Nova Network without tagging, Cinder LVM
2. Add 3 controllers, 2 compute nodes
3. Deploy cluster - deployment is successful
4. Assert that first controller is registered in nailgun - on this step we compare list with node's interfaces that we get after node creation by devops and list with interfaces that we get from http://10.108.0.2:8000/api/nodes
Expected result - dictionary with node information is returned
Actual - None is returned that means that slave wasn't found in nailgun
This issue periodically happens on CI
Logs are attached
Cobbler log
Tue Apr 15 22:44:19 2014 - INFO | Exception value: 'internal error, unknown distro name bootstrap'
Without bootstrap image provided it is not possible to register node, so nailgun correctly returns empty list