The aim of this project is to use Ansible role to export AWX Tower logs on disk.


Accessing log files isn’t always easy. Searching for errors across multiple log files, for workflows for example, or nested playbooks can be a difficult task.
The proposed role extracts logs between two dates, or between two id executions, and saves them to disk.
The role saves in parallel with the ansible execution log, the json file with each job’s description, name, parameters and all execution-related information.


The role is freely distributed from Github:


The variables used are described in the playbooks/vars/awx_logs_keeper_vars.yml as follows:

log_path_root: "/foo/bar/awx_logs"
  tower_base_url: ""
  tower_username: "{{ awx_username }}"
  tower_password: "{{ awx_password }}"
   start: 2
   end: 5
  start: "2022-10-20T10:00:00"
  end: "2022-10-20T13:00:00"
  • log_pat_root: Is the path where the logs will be saved
  • awx section is used for awx credentials and web sevrver address
  • job_id_interval
  • instruct the playbook to get logs got all jobs from job_id_interval.start and job_id_interval.end.
  • date_interval
  • instruct the playbook to get logs got all jobs between date_interval.start date and date_interval.end date.
  • date_interval.start must be lower than date_interval.end.
  • job_id_interval.start must be lower than job_id_interval.end.
  • Values of job_id_interval and date_interval are not mutual exclusive, but at least one of date_interval or job_id-interval must be given.

The playbook will test first if date_interval make sense, then it gets all job_id between these two dates. At the end it will add the job_id_interval, and get all logs for the obtained job_id list.

The output is written in log_path_root/date_of_job_execution/job_id.

Two files are produced :

  • output_<job_id>_<DDMMYYY_HHMISS>.log is the ansible output of the log.
  • <job_id>_<DDMMYYY_HHMISS>.json is the json description of the job.


# Execute the playbook by modifying the default value of log_path_root
$>  ansible-playbook playbooks/awx_logs_keeper_vars.yml -e "log_path_root=/tmp/awx_logs" --vault-password-file playbooks/vars/vault-pass

TASK [awx_get_one_log : Out the logfile path] **********************************************************************************************************
ok: [localhost] => {
    "msg": "Write Log for job id 324299 to /tmp/awx_logs/20-10-2022/324299"

$> cd /tmp/awx_logs/20-10-2022/324299
$> ls
324299_20102022_111602.json   output_324299_20102022_111602.log


In the example the --vault-password-file playbooks/vars/vault-pass is given to shaddow the Ansible Tower access credentials.

gather_facts: true
    - vars/secret.yml
    - vars/awx_logs_keeper_vars.yml
Shadowed parameters are:
* `awx_password`
* `awx_password`

Final thoughts

The log_keeper role can be scheduled on a daily basis, to give access to the previous day’s execution logs.

This makes error investigation much easier, especially for complex playbooks.

Share on