I have seen how to register variables within tasks in an ansible playbook and then use those variables elsewhere in the same playbook, but can you register a variable in an included playbook and then access those variables back in the original playbook?
Here is what I am trying to accomplish:
This is my main playbook:
- include: sub-playbook.yml job_url="http://some-jenkins-job"
- hosts: localhost
roles:
- some_role
sub-playbook.yml:
---
- hosts: localhost
tasks:
- name: Collect info from Jenkins Job
script: whatever.py --url "{{ job_url }}"
register: jenkins_artifacts
I'd like to be able to access the jenkins_artifacts results back in main_playbook if possible. I know you can access it from other hosts in the same playbook like this: "{{ hostvars['localhost']['jenkins_artifacts'].stdout_lines }}"
Is it the same idea for sharing across playbooks?
I'm confused what this question is about. Just use the variable name jenkins_artifacts:
- include: sub-playbook.yml job_url="http://some-jenkins-job"
- hosts: localhost
debug:
var: jenkins_artifacts
This might seem complicated but I love doing this in my Playbooks:
rc defines the name of the variable which contains the return value
ar gives the arguments to the include tasks
master.yml:
- name: verify_os
include_tasks: "verify_os/main.yml"
vars:
verify_os:
rc: "isos_present"
ar:
image: "{{ os.ar.to_os }}"
verify_os/main.yml:
---
- name: check image on device
ios_command:
commands:
- "sh bootflash: | inc {{ verify_os.ar.image }}"
register: image_check
- name: check if available
shell: "printf '{{ image_check.stdout_lines[0][0] }}\n' | grep {{ verify_os.ar.image }} | wc -l"
register: image_available
delegate_to: localhost
- set_fact: { "{{ verify_os.rc }}": "{{ true if image_available.stdout == '1' else false }}" }
...
I can now use the isos_present variable anywhere in the master.yml to access the returned value.
Related
I have 3 types of servers: dev, qa and prod. I need to send files to server specific home directories, i.e:
Dev Home Directory: /dev/home
QA Home Directory:/qa/home
PROD Home Directory: /prod/home
I have var set as boolean to determine the server type and think of using set_fact with condition to assign home directories for the servers. My playbook looks like this:
---
- hosts: localhost
var:
dev: "{{ True if <hostname matches dev> | else False }}"
qa: "{{ True if <hostname matches qa> | else False }}"
prod: "{{ True if <hostname matches prod> | else False }}"
tasks:
- set_facts:
home_dir: "{{'/dev/home/' if dev | '/qa/home' if qa | default('prod')}}"
However, then I ran the playbook, I was getting error about 'template expected token 'name', got string> Anyone know what I did wrong? Thanks!
Use match test. For example, the playbook
shell> cat pb.yml
- hosts: localhost
vars:
dev_pattern: '^dev_.+$'
qa_pattern: '^qa_.+$'
prod_pattern: '^prod_.+$'
dev: "{{ hostname is match(dev_pattern) }}"
qa: "{{ hostname is match(qa_pattern) }}"
prod: "{{ hostname is match(prod_pattern) }}"
tasks:
- set_fact:
home_dir: /prod/home
- set_fact:
home_dir: /dev/home
when: dev|bool
- set_fact:
home_dir: /qa/home
when: qa|bool
- debug:
var: home_dir
gives (abridged)
shell> ansible-playbook pb.yml -e hostname=dev_007
home_dir: /dev/home
Notes:
The variable prod is not used because /prod/home is default.
prod/home is assigned to home_dir first because it's the default. Next tasks can conditionally overwrite home_dirs.
Without the variable hostname defined the playbook will crash.
A simpler solution that gives the same result is creating a dictionary of 'pattern: home_dir'. For example
- hosts: localhost
vars:
home_dirs:
dev_: /dev/home
qa_: /qa/home
prod_: /prod/home
tasks:
- set_fact:
home_dir: "{{ home_dirs|
dict2items|
selectattr('key', 'in' , hostname)|
map(attribute='value')|
list|
first|default('/prod/home') }}"
- debug:
var: home_dir
Adding an alternative method to achieve this. This kind of extends the group_vars suggestion given by #mdaniel in his comment.
Ansible has a ready-made mechanism to build the variables based on the inventory hosts and groups. If you organize your inventory, you can avoid a lot of complication in trying to match host patterns.
Below is a simplified example, please go through the link above for more options.
Consider an inventory file /home/user/ansible/hosts:
[dev]
srv01.dev.example
srv02.dev.example
[qa]
srv01.qa.example
srv02.qa.example
[prod]
srv01.prod.example
srv02.prod.example
Using group_vars:
Then you can have below group_var files in /home/user/ansible/group_vars/ (matching inventory group names):
dev.yml
qa.yml
prod.yml
In dev.yml:
home_dir: "/dev/home"
In qa.yml:
home_dir: "/qa/home"
In prod.yml:
home_dir: "/prod/home"
Using host_vars:
Or you can have variables specific to hosts in host_vars directory /home/user/ansible/host_vars/:
srv01.dev.example.yml
srv01.prod.example.yml
# and so on
In srv01.dev.example.yml:
home_dir: "/dev/home"
In srv01.prod.example.yml:
home_dir: "/prod/home"
These variables will be picked based on which hosts you run the playbook, for example the below playbook:
---
- hosts: dev
tasks:
- debug:
var: home_dir
# will be "/dev/home"
- hosts: prod
tasks:
- debug:
var: home_dir
# will be "/prod/home"
- hosts: srv01.dev.example
tasks:
- debug:
var: home_dir
# will be "/dev/home"
- hosts: srv01.prod.example
tasks:
- debug:
var: home_dir
# will be "/prod/home"
For me the below code is working -
with_items: "{{ groups['mlpoc'] }}" but instead of hardcoded mlpoc I want to pass it in a variable as a parameter.
Say the command parameter is mlhosts=mlpoc and I want to use the variable instead of hardcoded value, something like -
with_items: "{{ groups['{{ mlhosts }}'] }}" but it throws error. Is it even possible to pass a dynamic value to groups ?
Here is my code -
- hosts: dbsrd3510
user: '{{ mluser }}'
gather_facts: no
no_log: false
tasks:
- name: Fetch source list from clients
with_items: "{{ groups['mlpoc'] }}"
shell: rsync -av /MLbackup/{{ pkg }} {{ mluser }}#{{ item }}:/tmp/
Try something like this,
---
- hosts: all
gather_facts: no
tasks:
- name: Add a line to a file if the file does not exist, without passing regexp
debug:
msg: "{{ item }}"
with_items: " {{ groups[group_name] }} "
And you can test the above changes using,
ansible-playbook -i hosts main.yml -e group_name="all"
I'm setting fact in a role:
- name: Check if manager already configured
shell: >
docker info | perl -ne 'print "$1" if /Swarm: (\w+)/'
register: swarm_status
- name: Init cluster
shell: >-
docker swarm init
--advertise-addr "{{ ansible_default_ipv4.address }}"
when: "'active' not in swarm_status.stdout_lines"
- name: Get worker token
shell: docker swarm join-token -q worker
register: worker_token_result
- set_fact:
worker_token: "{{ worker_token_result.stdout }}"
Then I want to access worker_token on another hosts. Here's my main playbook, the fact is defined in the swarm-master role
- hosts: swarm_cluster
become: yes
roles:
- docker
- hosts: swarm_cluster:&manager
become: yes
roles:
- swarm-master
- hosts: swarm_cluster:&node
become: yes
tasks:
- debug:
msg: "{{ worker_token }}"
I'm getting undefined variable. How to make it visible globally?
Of course it works perfectly if I run debug on the same host.
if your goal is just to access worker_token from on another host, you can use hostvars variable and iterate through the group where you've defined your variable like this:
- hosts: swarm_cluster:&node
tasks:
- debug:
msg: "{{ hostvars[item]['worker_token'] }}"
with_items: "{{ groups['manager'] }}"
If your goal is to define the variable globally, you can add a step to define a variable on all hosts like this:
- hosts: all
tasks:
- set_fact:
worker_token_global: "{{ hostvars[item]['worker_token'] }}"
with_items: "{{ groups['manager'] }}"
- hosts: swarm_cluster:&node
tasks:
- debug:
var: worker_token_global
I am trying to create a playbook to find out on which openstack server vm is running on. I have created a list of openstack servers in vars and used delegate_to with with_items to iterate through until find vm. I am using wc -l at the end of command and 1 will be success. The aim is, once os-server is found, store servername into a var so this can be used for rest of tasks in playbook. I am unable to get the os server name in a var from the list. I am not an ansible expert. Can anyone help to achieve this? Thanks
- hosts: localhost
vars:
openstack:
- reg1
- reg2
- reg3
- reg4
tasks:
- name: Command to find os server where vm exists
shell: somecommand-to-check-if-vm-exist | wc -l
delegate_to: "{{ item }}"
with_items: "{{ openstack }}"
register: found_server
retries: 1
delay: 1
until: found_server.stdout != "1"
- debug: var=found_server
- name: set fact
set_fact: os-server = "{{ item.item }}"
when: item.stdout == "1"
with_items: "{{ found_server.results }}"
register: var2
- name: debug var
debug: var=var2
- debug: var=os-server
There's no need to retry/until here and for the second loop as well.
Try this:
- hosts: localhost
vars:
openstack: [reg1, reg2, reg3, reg4]
tasks:
- name: Command to find os server where vm exists
shell: somecommand-to-check-if-vm-exist | wc -l
delegate_to: "{{ item }}"
with_items: "{{ openstack }}"
register: vm_check
- name: set fact
set_fact:
os_server: "{{ (vm_check.results | selectattr('stdout','equalto','1') | list | first).item }}"
- name: debug var
debug:
msg: "{{ os_server }}"
This will register results from every server into vm_check.results, and then just select elements with stdout set to 1, take first element of it it (I suppose you always have one server with VM), and get .item of this element which contains the item of original loop (in our case it is server's name).
We need to go through this structure
Zone spec
https://gist.github.com/git001/9230f041aaa34d22ec82eb17d444550c
I was able to run the following snipplet but now I'm stucked at the error checking.
playbook
--
- hosts: all
gather_facts: no
vars_files:
- "../doc/application-zone-spec.yml"
roles:
- { role: ingress_add, customers: "{{ application_zone_spec }}" }
role
- name: check if router exists
shell: "oc get dc -n default {{ customers.zone_name }}-{{ item.type }}"
with_items: "{{ customers.ingress }}"
ignore_errors: True
register: check_router
- name: Print ingress hostnames
debug: var=check_router
- name: create new router
shell: "echo 'I will create a router'"
with_items: "{{ customers.ingress }}"
when: check_router.rc == 1
Output of a ansible run
https://gist.github.com/git001/dab97d7d12a53edfcf2a69647ad543b7
The problem is that I need to go through the ingress items and I need to map the error of the differnt types from the "check_router" register.
It would be nice to make something like.
Pseudo code.
Iterate through the "customers.ingress"
check in "check_router" if the rc is ! 0
execute command.
We use.
ansible-playbook --version
ansible-playbook 2.1.0.0
config file = /etc/ansible/ansible.cfg
configured module search path = Default w/o overrides
You can replace the second loop with:
- name: create new router
shell: "echo 'I will create a router with type {{ item.item }}'"
with_items: "{{ check_router.results }}"
when: item.rc == 1
This will iterate over every step of check_route loop and you can access original items via item.item.