Conditional Execution Based on Shell Output
One of the most powerful capabilities in Ansible is the ability to make decisions based on the output of shell commands. In this step, we'll learn how to use conditions and filters to process shell command output and make playbooks more dynamic.
Using Conditions with Shell Output
Let's create a playbook that makes decisions based on shell command output:
- Create a new file called
conditional_playbook.yml in the /home/labex/project directory.
- Add the following content:
---
- name: Conditional Tasks Based on Command Output
hosts: local
gather_facts: no
tasks:
- name: Check disk space
shell: df -h / | grep -v Filesystem | awk '{print $5}' | sed 's/%//'
register: disk_usage
- name: Display disk usage
debug:
msg: "Current disk usage: {{ disk_usage.stdout }}%"
- name: Disk usage warning
debug:
msg: "WARNING: Disk usage is high"
when: disk_usage.stdout|int > 50
- name: Disk usage normal
debug:
msg: "Disk usage is normal"
when: disk_usage.stdout|int <= 50
This playbook:
- Runs a shell command to check disk usage percentage on the root filesystem
- Uses the
when conditional based on the command output
- Uses the
int filter to convert the string output to an integer for comparison
Run the playbook:
ansible-playbook -i inventory.ini conditional_playbook.yml
The output will vary depending on your actual disk usage, but will look something like this:
PLAY [Conditional Tasks Based on Command Output] ******************************
TASK [Check disk space] *******************************************************
changed: [localhost]
TASK [Display disk usage] *****************************************************
ok: [localhost] => {
"msg": "Current disk usage: 38%"
}
TASK [Disk usage warning] *****************************************************
skipped: [localhost]
TASK [Disk usage normal] ******************************************************
ok: [localhost] => {
"msg": "Disk usage is normal"
}
PLAY RECAP ********************************************************************
localhost : ok=3 changed=1 unreachable=0 failed=0 skipped=1 rescued=0 ignored=0
Notice how Ansible executed only one of the conditional tasks based on the actual disk usage value.
Handling JSON Output from Commands
Many modern CLI tools return data in JSON format. Ansible has built-in capabilities to handle JSON output:
- Create a new file called
json_output.yml in the /home/labex/project directory.
- Add the following content:
---
- name: Handling JSON Output
hosts: local
gather_facts: no
tasks:
- name: Create a JSON file for testing
copy:
dest: /tmp/services.json
content: |
{
"services": [
{
"name": "web",
"status": "running",
"port": 80
},
{
"name": "database",
"status": "stopped",
"port": 5432
},
{
"name": "cache",
"status": "running",
"port": 6379
}
]
}
- name: Read JSON file with shell
shell: cat /tmp/services.json
register: json_output
- name: Parse and display JSON content
debug:
msg: "{{ json_output.stdout | from_json }}"
- name: Extract and display service information
debug:
msg: "Service: {{ item.name }}, Status: {{ item.status }}, Port: {{ item.port }}"
loop: "{{ (json_output.stdout | from_json).services }}"
- name: Show only running services
debug:
msg: "Running service: {{ item.name }} on port {{ item.port }}"
loop: "{{ (json_output.stdout | from_json).services }}"
when: item.status == "running"
This playbook:
- Creates a sample JSON file for demonstration
- Reads the JSON file with a shell command
- Uses the
from_json filter to parse the JSON string into a data structure
- Loops through the data structure to display information
- Uses conditional logic to filter only running services
Run the playbook:
ansible-playbook -i inventory.ini json_output.yml
You should see output similar to this:
PLAY [Handling JSON Output] ***************************************************
TASK [Create a JSON file for testing] *****************************************
changed: [localhost]
TASK [Read JSON file with shell] **********************************************
changed: [localhost]
TASK [Parse and display JSON content] *****************************************
ok: [localhost] => {
"msg": {
"services": [
{
"name": "web",
"port": 80,
"status": "running"
},
{
"name": "database",
"port": 5432,
"status": "stopped"
},
{
"name": "cache",
"port": 6379,
"status": "running"
}
]
}
}
TASK [Extract and display service information] ********************************
ok: [localhost] => (item={'name': 'web', 'status': 'running', 'port': 80}) => {
"msg": "Service: web, Status: running, Port: 80"
}
ok: [localhost] => (item={'name': 'database', 'status': 'stopped', 'port': 5432}) => {
"msg": "Service: database, Status: stopped, Port: 5432"
}
ok: [localhost] => (item={'name': 'cache', 'status': 'running', 'port': 6379}) => {
"msg": "Service: cache, Status: running, Port: 6379"
}
TASK [Show only running services] *********************************************
ok: [localhost] => (item={'name': 'web', 'status': 'running', 'port': 80}) => {
"msg": "Running service: web on port 80"
}
skipped: [localhost] => (item={'name': 'database', 'status': 'stopped', 'port': 5432})
ok: [localhost] => (item={'name': 'cache', 'status': 'running', 'port': 6379}) => {
"msg": "Running service: cache on port 6379"
}
PLAY RECAP ********************************************************************
localhost : ok=5 changed=2 unreachable=0 failed=0 skipped=0 rescued=0 ignored=0
Notice how the playbook parses the JSON, extracts specific information, and filters the data based on conditions.
Error Handling with Shell Commands
When running shell commands, it's important to handle potential errors gracefully:
- Create a new file called
error_handling.yml in the /home/labex/project directory.
- Add the following content:
---
- name: Error Handling with Shell Commands
hosts: local
gather_facts: no
tasks:
- name: Run a potentially failing command
shell: grep "nonexistent_pattern" /etc/passwd
register: command_result
ignore_errors: yes
- name: Display success or failure
debug:
msg: "Command {{ 'succeeded' if command_result.rc == 0 else 'failed with return code ' + command_result.rc|string }}"
- name: Run a custom failing command
shell: exit 3
register: exit_command
ignore_errors: yes
- name: Display detailed error information
debug:
msg: |
Return code: {{ exit_command.rc }}
Error message: {{ exit_command.stderr if exit_command.stderr else 'No error message' }}
This playbook:
- Runs commands that are expected to fail
- Uses
ignore_errors: yes to continue the playbook execution even when commands fail
- Shows different methods for handling and displaying error information
Run the playbook:
ansible-playbook -i inventory.ini error_handling.yml
You should see output similar to this:
PLAY [Error Handling with Shell Commands] *************************************
TASK [Run a potentially failing command] **************************************
fatal: [localhost]: FAILED! => {"changed": true, "cmd": "grep \"nonexistent_pattern\" /etc/passwd", "delta": "0:00:00.002916", "end": "2023-07-14 16:10:23.671519", "msg": "non-zero return code", "rc": 1, "start": "2023-07-14 16:10:23.668603", "stderr": "", "stderr_lines": [], "stdout": "", "stdout_lines": []}
...ignoring
TASK [Display success or failure] *********************************************
ok: [localhost] => {
"msg": "Command failed with return code 1"
}
TASK [Run a custom failing command] *******************************************
fatal: [localhost]: FAILED! => {"changed": true, "cmd": "exit 3", "delta": "0:00:00.002447", "end": "2023-07-14 16:10:23.906121", "msg": "non-zero return code", "rc": 3, "start": "2023-07-14 16:10:23.903674", "stderr": "", "stderr_lines": [], "stdout": "", "stdout_lines": []}
...ignoring
TASK [Display detailed error information] *************************************
ok: [localhost] => {
"msg": "Return code: 3\nError message: No error message\n"
}
PLAY RECAP ********************************************************************
localhost : ok=3 changed=2 unreachable=0 failed=0 skipped=0 rescued=0 ignored=2
This demonstrates how to capture and respond to different error conditions when running shell commands.
