tdp deploy doesn't update database

[giem-git]

Description

• tdp_lib database is not updated on fail.
• It also looks like the database is not updated immediately for success operations. (a select in database a long time after success shows first operations status still 'PLANNED'
• This behavior appeared in commit cde5e14

Reproducing :

• break the first planned role (in my configuration, I change indentation in tdp/ansible/ansible_collections/tosit/tdp/roles/exporter/jmx/tasks/install.yml)
• cleanup : rm sqlite.db && tdp init
• deploy : we expect this to fail

  (tdp-ops) > $ tdp plan dag && tdp deploy
  2023-10-27 10:16:57,383 - INFO - root - Logging is configured.
  Creating a deployment plan for the whole DAG.
  Deployment plan successfully created.
  2023-10-27 10:16:58,289 - INFO - root - Logging is configured.
  ERROR! We were unable to read either as JSON nor YAML, these are the errors we got from each:
  JSON: Expecting value: line 1 column 1 (char 0)
  
  Syntax Error while loading YAML.
    mapping values are not allowed in this context
  
  The error appears to be in 'xxx/dev/tdp-ops/tdp/ansible/ansible_collections/tosit/tdp/roles/exporter/jmx/tasks/install.yml': line 23, column 25, but may
  be elsewhere in the file depending on the exact syntax problem.
  
  The offending line appears to be:
  
  - name: Create symbolic link to jmx exporter jar
      ansible.builtin.file:
                          ^ here
  
  

• browse logs :

  (tdp-ops) > $ tdp browse 
  2023-10-27 10:20:16,894 - INFO - root - Logging is configured.
    id  start_time           end_time    status    deployment_type
  ----  -------------------  ----------  --------  -----------------
     1  2023-10-27 10:16:58              RUNNING   DAG
                                                                                                                                                                                                      
  (tdp-ops) > $ tdp browse 1 |head -20
  2023-10-27 10:21:03,933 - INFO - root - Logging is configured.
  Deployment details
  id               1
  options          {}
  start_time       2023-10-27 10:16:58
  end_time
  status           RUNNING
  deployment_type  DAG
  
  
  Operations
    deployment_id    operation_order  operation                                 host    extra_vars    start_time    end_time    state    logs
  ---------------  -----------------  ----------------------------------------  ------  ------------  ------------  ----------  -------  ------
                1                  1  exporter_jmx_install                                                                      PLANNED
                1                  2  exporter_node_install                                                                     PLANNED
                1                  3  exporter_postgres_install                                                                 PLANNED
                1                  4  exporter_install                                                                          PLANNED
                1                  5  zookeeper_client_install                                                                  PLANNED
                1                  6  zookeeper_server_install                                                                  PLANNED
                1                  7  zookeeper_kerberos_install                                                                PLANNED
                1                  8  zookeeper_install                                                                         PLANNED
  
  

As a result, tdp resume fails with following error : tdp.core.models.deployment_log.NothingToResumeError: Nothing to resume, deployment #1 was Running.

[PaulFarault]

Thanks for the issue. I'll have a look at it!

[PaulFarault]

This should have been fixed by #499 and #500. Would you be able to check please?

[giem-git]

This should have been fixed by #499 and #500. Would you be able to check please?

Issue is solved. Thanks !