How to use return values of one task in another task for a different host in ansible

Question

I was trying to setup mysql master slave replication with ansible for a hostgroup consisting of 2 mysql hosts.

Here is my scenario:

I run one task in the 1st host and skips the 2nd host, so the 1st task (i.e master replication status) returns some values like Position, File etc.

Then, I run another task in 2nd host (skips the 1st hosts), This task uses the return values of the 1st task like master.Position, master.File etc.

Now, when I run the playbook, the variables of the 1st task does not seem to be working in the 2nd task

Inventory File

    [mysql]
    stagmysql01 ansible_host=1.1.1.1 ansible_ssh_user=ansible ansible_connection=ssh
    stagmysql02 ansible_host=1.1.1.2 ansible_ssh_user=ansible ansible_connection=ssh

Tasks on Master

- name: Mysql - Check master replication status.
  mysql_replication: mode=getmaster
  register: master

- debug: var=master

Tasks on Slave

- name: Mysql - Configure replication on the slave.
  mysql_replication:
    mode: changemaster
    master_host: "{{ replication_master }}"
    master_user: "{{ replication_user }}"
    master_password: "{{ replication_pass }}"
    master_log_file: "{{ master.File }}"
    master_log_pos: "{{ master.Position }}"
  ignore_errors: True

Master Output

TASK [Mysql_Base : Mysql - Check master replication status.] ****************
skipping: [stagmysql02]
ok: [stagmysql01]

TASK [Mysql_Base : debug] ***************************************************
ok: [stagmysql01] => {
    "master": {
        "Binlog_Do_DB": "", 
        "Binlog_Ignore_DB": "mysql,performance_schema", 
        "Executed_Gtid_Set": "", 
        "File": "mysql-bin.000003", 
        "Is_Master": true, 
        "Position": 64687163, 
        "changed": false, 
        "failed": false
    }
}
ok: [stagmysql02] => {
    "master": {
        "changed": false, 
        "skip_reason": "Conditional result was False", 
        "skipped": true
    }
}

Slave Output

TASK [Mysql_Base : Mysql - Configure replication on the slave.] *************
skipping: [stagmysql01]
fatal: [stagmysql02]: FAILED! => {"failed": true, "msg": "The task includes an option with an undefined variable. The error was: 'dict object' has no attribute 'File'\n\nThe error appears to have been in '/root/ansible/roles/Mysql_Base/tasks/replication.yml': line 30, column 3, but may\nbe elsewhere in the file depending on the exact syntax problem.\n\nThe offending line appears to be:\n\n\n- name: Mysql - Configure replication on the slave.\n  ^ here\n\nexception type: <class 'ansible.errors.AnsibleUndefinedVariable'>\nexception: 'dict object' has no attribute 'File'"}
...ignoring

As, you can see above, the 2nd task failed for 2nd host because of undefined variables. However the required variables are there in 1st task of 1st host.

How do I use the variables returned from 1st host in 2nd host in another task ?

P.S: I have seen the approach of using {{ hostvars['inventory_hostname']['variable'] }}. However I'm quite confused with this approach as the inventory_hostname or IP address needs to be added directly. I was looking for a common template that can be used for different inventory files and playbooks.

So what is the question in that case? You set the inventory name and you refer to it in the playbook. Problem solved. It’s not an IP address or a hostname which you might not know in advance. — techraf
If have different environments like , staging, production etc. and I have environment specific inventory files and playbooks, I want to use the same role for all environments. So the tasks in the roles cannot refer a inventory name in particular, as it would be different for other inventory files. — Jerald Sabu M
I’m sorry, but what you wrote makes no sense to me. Ansible was designed exactly to solve this problem by using inventories. Just start with tutorials on docs.ansible.com — techraf
In the question, you can see that I run a task in the 1st host, skips the 2nd hosts and task returns values for variables like Position, File etc only for 1st host. I run the another task in the 2nd host, this time 1st host is skipped. How to use the variables of 1st host in 2nd host ? — Jerald Sabu M
You have included the proper syntax in the question above. Other concerns constitute an XY Problem — techraf

Jerald Sabu M Jerald Sabu M · Accepted Answer · 2018-01-01T17:05:58

I was able to solve my problem by defining the variables to a new dummy host and then using it across the playbook with hostvars.

Similar solution was already mentioned in one of the answers in How do I set register a variable to persist between plays in ansible? However I did not notice it until I posted this question.

Here is what I did in the ansible tasks:

I have created a dummy host master_value_holder and defined the required variables. (Here I needed master_log_file and master_log_Postion)
Accessed the variables using hostvars['master_value_holder']['master_log_file']

Tasks on Master

- name: Mysql - Check master replication status.
  mysql_replication: mode=getmaster
  register: master

- name: "Add master return values to a dummy host"
  add_host:
    name:   "master_value_holder"
    master_log_file: "{{ master.File }}"
    master_log_pos: "{{ master.Position }}"

Tasks for Slave

- name: Mysql - Displaying master replication status
  debug: msg="Master Bin Log File  is {{ hostvars['master_value_holder']['master_log_file'] }} and Master Bin Log Position is {{ hostvars['master_value_holder']['master_log_pos'] }}"

- name: Mysql - Configure replication on the slave.
  mysql_replication:
    mode: changemaster
    master_host: "{{ replication_master }}"
    master_user: "{{ replication_user }}"
    master_password: "{{ replication_pass }}"
    master_log_file: "{{ hostvars['master_value_holder']['master_log_file'] }}"
    master_log_pos: "{{ hostvars['master_value_holder']['master_log_pos'] }}"
  when: ansible_eth0.ipv4.address != replication_master and not slave.Slave_SQL_Running

Output

TASK [Mysql_Base : Mysql - Check master replication status.] ****************
skipping: [stagmysql02]
ok: [stagmysql01]

TASK [AZ-Mysql_Base : Add master return values to a dummy host] ****************
changed: [stagmysql01]

TASK [AZ-Mysql_Base : Mysql - Displaying master replication status] ************
ok: [stagmysql01] => {
    "msg": "Master Bin Log File  is mysql-bin.000001 and Master Bin Log Position is 154"
}
ok: [stagmysql02] => {
    "msg": "Master Bin Log File  is mysql-bin.000001 and Master Bin Log Position is 154"
}

TASK [AZ-Mysql_Base : Mysql - Configure replication on the slave.] *************
skipping: [stagmysql01]
skipping: [stagmysql02]

As you can see from the above output that the master replication status is available for both the hosts now.

How to use return values of one task in another task for a different host in ansible

1 Answers