distcloud/distributedcloud/dcmanager/orchestrator/states
Victor Romano f081453b3c Add subcloud deploy abort option to dcmanager
This commit adds the command "subcloud deploy abort" to dcmanager.
It allows the user to abort the on-going execution of a playbook
against the subcloud. Any task will be aborted immediately if the
unabortable flag is not set by the playbook [1]. If current
operation is install, a shutdown command will also be issued.

Test Plan:
  Success cases:
  - PASS: Trigger an abort during installation and verify that the
          playbook execution was aborted immediately, the subcloud was
          shut dows after it and the RVMC pod and job were terminated.
  - PASS: Trigger an abort during config and verify that
          the playbook execution was aborted immediately.
  - PASS: Trigger an abort during bootstrap without the presence
          of unabortable flag and verify that the playbook execution
          was aborted immidiately.
  - PASS: Trigger an abort during bootstrap with the presence of
          unabortable flag and verify that the playbook execution
          was aborted only after the flag was deleted.
  - PASS: Trigger an abort directly calling the API (using CURL
          instead of using the CLI.
  Failure cases:
  - PASS: Verify that the abort request is rejected if deploy
          state is not 'installing', 'bootstrapping' or 'configuring'.
  - PASS: Abort when an unabortable task is running and then force an
          external error during this task restarting dcmanager-manager
          service, verify that deploy state is set to a failed state
          e.g. bootstrap-failed.
  - PASS: Abort when an unabortable task is running and then force an
          internal error during this task using ansible.builtin.fail
          module, verify that deploy state is set to a failed state
          e.g. bootstrap-failed.
  - PASS: Abort when an unabortable task is running and then force
          the playbook to halt the execution during this task using
          ansible.builtin.pause module, verify that deploy state is
          set to a failed state after 10 minutes
          e.g. bootstrap-failed.

Story: 2010756
Task: 48102

Co-Authored-By: Gustavo Herzmann <gustavo.herzmann@windriver.com>
Signed-off-by: Victor Romano <victor.gluzromano@windriver.com>
Change-Id: Ic5311324a76bf7ce1215692e934d5577ff82868e
2023-07-19 14:16:00 -03:00
..
firmware Tech debt: move portions of dcorch.common.consts into dccommon.consts 2022-06-16 12:28:44 -04:00
kube Ignore certain alarms during kube upgrade pre-check 2023-05-04 18:00:26 -03:00
kube_rootca Increment timeout of DC K8s Root CA strategy 2023-03-03 11:39:31 -03:00
patch Add the upload-only option to DC patch orchestration 2023-03-02 13:56:18 -03:00
prestage Add release optionality to subcloud prestage 2023-05-19 16:33:42 -04:00
upgrade Add subcloud deploy abort option to dcmanager 2023-07-19 14:16:00 -03:00
__init__.py Move dcmanager orchestration to a separate process 2020-09-14 11:17:06 -04:00
applying_vim_strategy.py Increment timeout of DC K8s Root CA strategy 2023-03-03 11:39:31 -03:00
base.py Refactor distributed cloud patch orchestration 2023-02-27 09:48:14 -03:00
creating_vim_strategy.py Distributed Cloud Kubernetes Upgrade Orchestration 2021-03-03 11:24:14 -06:00
lock_host.py Upgrade Orchestration for Standard subclouds 2021-01-28 05:25:54 -05:00
swact_host.py Raise orchestrated strategy post-swact wait time 2022-06-13 18:07:18 -03:00
unlock_host.py Upgrade orchestration updates for virtual subclouds 2022-06-17 22:45:06 -04:00