Community Training Classes & Labs > F5 Programmability Training Index

Appendix A: Python SDK

This module will cover the newly released F5 Python SDK. This SDK is released and maintained as a public GitHub repository at

The goal of the Python SDK is to provide a simple interface that abstracts many of the F5-specific nuances of the iControl REST API away from the user. As you learned in Module 1, when interacting directly with the API, it’s often necessary to build out requests in a very manual fashion. In order to provide a simpler interface, the SDK was developed to abstract away many of the eccentricities of the API and provide a clean, Pythonic interface.

For example, when creating a pool in, an Imperative automation model, without the SDK you would be required to do something like the following (this code is not complete):

import requests
import sys
base_url = “”

pool_attributes = {
        “name”: “test_pool”,
        “partition”: “Common”,
        “loadBalancingMode”: “least-connections-member”,
        “minUpMembers”: 1

s = requests.session()
s.auth = (“admin”, “admin”)

resp =, data=json.dumps(pool_attributes))

if resp.status_code !=
        print “Error creating pool”


When using the Python SDK the equivalent code is:

from f5.bigip import ManagementRoot

mgmt = ManagementRoot(“”,”admin”,”admin”)

pool =”Common”, name=”test_pool”)
pool.loadBalancingMode = “least-connections-member”
pool.minUpMembers = 1


As you can see, the code utilizing the SDK is much more condensed and far easier to read. This is a result of the SDK exposing abstracted methods to build the URL. Additionally the SDK creates standard CURDLE (create, update, refresh, delete, load, exists) methods that behave correctly depending on REST object type (Organizing Collection, Resource, etc.) you are interacting with (e.g., you cannot DELETE an Organizing Collection, therefore a delete() method is not available).

Full documentation for the API exists at here

For the purpose of this lab, your Windows Jumphost has everything pre-installed, however, since the SDK is a standard python package the process is trivial on any system (Windows, Linux, Mac, etc.) that has Python installed.

It’s important to keep in mind, while going through this module, that we are only demonstrating what is possible with the SDK from a high level. For example, the same scripts used in this module are designed to run from the command line with arguments, however, they could easily be modified to use JSON files as the input mechanism.