Basic example using nornir_scrapli and Cisco Always-On Sandbox

[writememe]

Overview

After a question in slack, I thought I'd make a very quick example so that people have something to work with..

Below is a cut-down basic example of using nornir_scrapli and the Cisco Always-On IOS-XE Sandbox

It performs the following:

• Connects to the sandbox host
• Runs a single command
• Prints output to screen

File Structure

There are four files that you need to generate/copy into your local machine:

.
├── basic_nornir_scrapli_example.py
├── groups.yaml
├── hosts.yaml
└── requirements.txt

requirements.txt

This file contains the python requirements to run this example. Ensure that they are on your machine/python virtual environment:

nornir
nornir-scrapli

hosts.yaml

This file contains the details for IOS-XE Sandbox host. In production, DO NOT save your passwords here in a clear-text, unencrypted file:

---
sandbox-iosxe-latest-1.cisco.com:
    # NOTE: Bad practice in Production, don't do this.
    username: developer
    password: C1sco12345
    hostname: sandbox-iosxe-latest-1.cisco.com
    groups:
        - iosxe

groups.yaml

This file contains a single group to setup the iosxe specific settings.

---
iosxe:
    platform: ios
    # Setting scrapli connection options
    # https://github.com/carlmontanari/nornir_scrapli#a-simple-example
    connection_options:
        scrapli:
            platform: cisco_iosxe
            port: 22
            extras:
                auth_strict_key: false

basic_nornir_scrapli_example.py

Finally, here is the script which performs the tasks:

# Import modules
from nornir import InitNornir
import os
from nornir_scrapli.tasks import send_command
from nornir.core.task import Task
from nornir.core.task import Result

# Get path of the current dir under which the file is executed
dirname = os.path.dirname(os.path.abspath(__file__))


def get_nr():
   """
   Initialises a Nornir inventory using various configuration files.
   """
   # Initialise nornir
   nr = InitNornir(
       inventory={
           "options": {
               "host_file": os.path.join(
                   dirname,
                   "hosts.yaml",
               ),
               "group_file": os.path.join(
                   dirname,
                   "groups.yaml",
               ),
           }
       }
   )
   return nr


def scrapli_command_runner(task: Task, command: str = "show run") -> Result:
   """Run a single command on a host using Scrapli.

   Args:
       task: The nornir task object.
   """
   # Printout some diagnostic information
   print(f"Attempting command: '{command}' on host: '{task.host}' ...")
   # Run command using scrapli
   cmd_output = task.run(
       task=send_command, name=f"Execute Command: {command}", command=command
   )
   # Assign result of scrapli output and return
   result = Result(host=task.host, result=cmd_output.result)
   return result


# Initialise inventory
nr = get_nr()
# Specify command to run
command = "show run"
# Execute scrapli command runner on entire inventory
cmd_result = nr.run(task=scrapli_command_runner, command=command)
# Iterate over hosts and results and print them out
for host in cmd_result:
   # Printout some diagnostic information
   print(f"Host: '{host}' - Command: '{command}' result:")
   # Printout a seperator
   print("=" * 20)
   # Printout the results of the command output
   print(f"{cmd_result[host].result}")
   # Printout a seperator
   print("=" * 20)

Ideas to extend the basic example

Here are some ideas on how you can extend this to learn more about scrapli or nornir:

• Save the outputs to a file
• Write another function which takes a list of commands
• Run commands on other supported drivers