I'm trying to fire-and-forget some tasks and collect results after that. Here's my playbook:
---
- hosts: node2
gather_facts: yes
tasks:
- name: 'Some long script no 1 on node2'
shell: "time sleep $[ ( $RANDOM % 20 ) + 20 ]s"
async: 40
poll: 0
register: script1
- name: 'Another long script no 2 on node2'
shell: "time sleep $[ ( $RANDOM % 20 ) + 20 ]s"
async: 40
poll: 0
register: script2
- hosts: node2
tasks:
- name: "Collect results"
async_status:
jid: loop_item.ansible_job_id
loop:
- script1
- script2
loop_control:
loop_var: loop_item
register: async_poll_results
until: async_poll_results.finished
retries: 30
When I run it, I receive following error:
PLAY [node2] ****************************************************************************************************************************
TASK [Gathering Facts] ******************************************************************************************************************
ok: [hostname]
TASK [Some long script no 1 on node2] ***************************************************************************************************
changed: [hostname] => {"ansible_job_id": "814448842231.125544", "changed": true, "finished": 0, "results_file": "/home/external.kamil.sacharczuk/.ansible_async/814448842231.125544", "started": 1}
TASK [Another long script no 2 on node2] ************************************************************************************************
changed: [hostname] => {"ansible_job_id": "586999441005.125616", "changed": true, "finished": 0, "results_file": "/home/external.kamil.sacharczuk/.ansible_async/586999441005.125616", "started": 1}
PLAY [node2] ****************************************************************************************************************************
TASK [Gathering Facts] ******************************************************************************************************************
ok: [hostname]
TASK [Collect results] ******************************************************************************************************************
failed: [hostname] (item=script1) => {"ansible_job_id": "loop_item.ansible_job_id", "attempts": 1, "changed": false, "finished": 1, "loop_item": "script1", "msg": "could not find job", "started": 1}
failed: [hostname] (item=script2) => {"ansible_job_id": "loop_item.ansible_job_id", "attempts": 1, "changed": false, "finished": 1, "loop_item": "script2", "msg": "could not find job", "started": 1}
to retry, use: --limit @xxxxxxxxx
PLAY RECAP ******************************************************************************************************************************
hostname : ok=4 changed=2 unreachable=0 failed=1
Don't really know why I receive this "could not find job". I tried to run this "collect" tasks localy first, than I figured out, that this job results are stored on node2, so I run it there. Tried with or without gathering facts. Tried also use
hostvars['hostname'][loop_item][ansible_job_id]
but this gave me same error as here.
Any help would be much appreciated!
PS. I am running ansible 2.6.1
Please try as below with proper quotation::
- hosts: localhost
gather_facts: false
tasks:
- name: 'Some long script no 1 on node2'
shell: "time sleep $[ ( $RANDOM % 20 ) + 20 ]s"
async: 40
poll: 0
register: script1
- name: 'Another long script no 2 on node2'
shell: "time sleep $[ ( $RANDOM % 20 ) + 20 ]s"
async: 40
poll: 0
register: script2
- hosts: localhost
tasks:
- name: "Collect results"
async_status:
jid: "{{ loop_item.ansible_job_id }}"
loop:
- "{{ script1 }}"
- "{{ script2 }}"
loop_control:
loop_var: loop_item
register: async_poll_results
until: async_poll_results.finished
retries: 30