After environment reset, subsequent deploy fails due to use of incorrect Cobbler template

[Alexander Vlasov]

Detailed bug description:

     After deploy successfully finished or failed after lets say error during some plugin installation and reseted, subsequent deploy fails.
After "Deploy" button pushed astute tries to reboot all nodes to bootstrap image (although they already booted with bootstrap image). Config files for PXE boot in cobbler container at this time contain links to bootstrap image. But by some reason eventually instead of cobbler bootstrap profile nodes try to boot Ubuntu_1404_x86_64 profile. As result we can observe situation on the screenshot https://drive.google.com/open?id=0BxGaSao6dmy6dUdPNFMwUW9KalE
At this time nodes stuck in that state. deploy fails by timeout. If just reboot node after deploy failed we can observe following picture https://drive.google.com/open?id=0BxGaSao6dmy6d3NRbWNHZnF1Sjg
So PXE boot configuration is change and node is trying to boot from local drive but as previous step failed it is not possible.
I have diagnostic snapshot but due to slow internet connection can't attach it right now. Will do this ASAP.

Steps to reproduce:

    1) Install Fuel 7.0
    2) Create env: CEPH for all, Murano, VXLANs;
    3) Deploy environment.
    4) Reste environment.
    5) Deploy again and watch it failing

Expected results:

    Second deploy attempt should succeed

Actual result:

    Deploy fails

Workaround:

    As a workaround it is possible delete environment and create it from scratch.

VERSION:
  feature_groups:
    - mirantis
  production: "docker"
  release: "7.0"
  openstack_version: "2015.1.0-7.0"
  api: "1.0"
  build_number: "301"
  build_id: "301"
  nailgun_sha: "4162b0c15adb425b37608c787944d1983f543aa8"
  python-fuelclient_sha: "486bde57cda1badb68f915f66c61b544108606f3"
  fuel-agent_sha: "50e90af6e3d560e9085ff71d2950cfbcca91af67"
  fuel-nailgun-agent_sha: "d7027952870a35db8dc52f185bb1158cdd3d1ebd"
  astute_sha: "6c5b73f93e24cc781c809db9159927655ced5012"
  fuel-library_sha: "5d50055aeca1dd0dc53b43825dc4c8f7780be9dd"
  fuel-ostf_sha: "2cd967dccd66cfc3a0abd6af9f31e5b4d150a11c"
  fuelmain_sha: "a65d453215edb0284a2e4761be7a156bb5627677"

Diagnostic snapshot:
https://drive.google.com/open?id=0BxGaSao6dmy6alltS3ZlcUdNcE0

[Anton Chevychalov (achevychalov)]

Unable to reproduce problem on test environment. Besides we have such scenario in CI.

Perhaps we have some additional steps or something that change flow. Do you have any plugins or custom changes?

[Alexander Vlasov (avlasov)]

On the plugins side there were:
 - lbaas
 - ldap
 - LMA
 - Network plugin(add vlan network to vxlan)

But how could plugins affect provision step?

[Stanislaw Bogatkin (sbogatkin)]

Closed as invalid due to inactivity more than one month.