Some devices, such routers (cisco) need interactive session, in reason sub-menu hierarchy. It's just modification of standard paramiko pluging, to keep session active and allow walk trough menu.
---
# simple playbook for testing
#- debug: var=result
#- debug: var=result.stdout_lines
#- debug: var=result.stdout.find("#")
- hosts: cisco
gather_facts: False
tasks:
- name: check "enable" mode (super user mode)
raw: ;empty command # it's just empty comment in console
register: check_mode
- name: 'enable'
raw: enable
when: check_mode.stdout.find("#") == -1
register: result # not used
- name: 'configure terminal'
raw: conf t
when: check_mode.stdout.find("#") != -1
register: result
failed_when: result.stdout.find("(config)") == -1
- name: 'interface configuration fa0/1'
raw: int fa0/1
register: check_mode
failed_when: check_mode.stdout.find("(config-if)") == -1
- name: no shutdown
raw: shutdown
register: result
- name: close connection
raw: exit
register: result
Project directory includes ready for test environment based on GNS3 simulator inside vagrant.
Everything should be ready after
vagrant up
After connecting to virtual machine
vagrant ssh
tmux loaded automatically with 3 tabs.
telnet 127.0.0.1 2501 # or telnet 127.0.0.1 2502 call from console
In those tabs it's easy to control what commands successfully executed on routers Output example:
*Mar 1 00:19:47.679: %HA_EM-6-LOG: CLIaccounting: interface FastEthernet0/1
*Mar 1 00:19:48.719: %HA_EM-6-LOG: CLIaccounting: description connection to R
*Mar 1 00:19:49.631: %HA_EM-6-LOG: CLIaccounting: interface FastEthernet0/0
*Mar 1 00:19:50.651: %HA_EM-6-LOG: CLIaccounting: description not use
*Mar 1 00:19:51.563: %HA_EM-6-LOG: CLIaccounting: shutdown
If command is unsuccessful or wrong syntax, it missing inside output.
To capture such errors failed_when: need inside playbook.
# direct call
cd /vagrant/play && source ansible/hacking/env-setup
ansible-playbook R1_configure.yml
and
ansible-playbook R2_configure.yml