ipfabric 6.3.0

Python package for interacting with IP Fabric

IP Fabric

IPFabric is a Python module for connecting to and communicating against an IP Fabric instance.

About

Founded in 2015, IP Fabric develops network infrastructure visibility and analytics solution to help enterprise network and security teams with network assurance and automation across multi-domain heterogeneous environments. From in-depth discovery, through graph visualization, to packet walks and complete network history, IP Fabric enables to confidently replace manual tasks necessary to handle growing network complexity driven by relentless digital transformation.

v6.3.1 Deprecation Notices

In ipfabric>=v6.3.1 Python 3.7 support will be removed. This was originally planned for v7.0.0 however to add new functionality of Pandas Dataframe we are required to move this forward.

Python 3.7 is now End of Life as of June 27th 2023

v7.0.0 Deprecation Notices

In ipfabric>=v7.0.0 the following will be deprecated:

• ipfabric_diagrams package will move to ipfabric.diagrams
• The use of token='<TOKEN>' or username='<USER>', password='<PASS>' in IPFClient() will be removed:
  • Token: IPFClient(auth='TOKEN')
  • User/Pass: IPFClient(auth=('USER', 'PASS'))
  • .env file will only accept IPF_TOKEN or (IPF_USERNAME and IPF_PASSWORD) and not auth

Versioning

Starting with IP Fabric version 5.0.x the python-ipfabric and python-ipfabric-diagrams will need to match your IP Fabric version. The API's are changing and instead of api/v1 they will now be api/v5.0.

Version 5.1 will have backwards compatability with version 5.0 however 6.0 will not support any 5.x versions. By ensuring that your ipfabric SDK's match your IP Fabric Major Version will ensure compatibility and will continue to work.

Streaming Data Support

In IP Fabric version 6.3.0 the option to return table data using a streaming GET request instead of a paginated POST request has been added. This will be defaulted to True in the next Minor bump of the SDK (v6.4.0 or v7.0.0).

FOR CUSTOMERS USING RBAC THIS IS NOT RECOMMENDED. A bug has been identified where custom RBAC Policies do not allow you to create a Policy to the GET endpoints and only admins can query data. This is to be fixed in IP Fabric 6.3.1. THIS AFFECTS CSV EXPORT AND STREAMING JSON EXPORT.

• GET URL is limited to 4096 characters, complex queries and filters could go over this limit; however in testing it was very difficult to reach this.
• Since request has been changed from httpx.post to httpx.stream no changes in timeout was required in testing.
• Performance Testing on 1.7M rows:
  • POST requires 1,719 requests (1k rows per request) ~ 82 minutes
  • Streaming GET requires 1 request ~ 6.2 minutes
• No degradation in navigating the GUI including viewing table data or creating diagrams.
• Supports csv and json exports:
  • CSV
    • Only supported with a streaming GET request and return a bytes string of data in the Python SDK.
    • It will also convert times to human-readable format.
    • reports (returning Intent Check data) is not supported with CSV export
  • JSON provides same support as POST.

from ipfabric import IPFClient
ipf = IPFClient(streaming=True)

dev = ipf.inventory.devices.all()
dev_2 = ipf.fetch_all('tables/inventory/devices')
print(dev == dev_2)  # True
print(type(dev))  # list 

dev_csv = ipf.inventory.devices.all(export='csv')
dev_csv_2 = ipf.fetch_all('tables/inventory/devices', export='csv')
print(dev_csv == dev_csv_2 ) # True
print(type(dev_csv))  # bytes 

# Timezone can be changed for CSV export; see `ipfabric.tools.shared.TIMEZONES`
dev_csv_tz = ipf.inventory.devices.all(export='csv', csv_tz='UTC')

# If specifying to return reports and CSV request will drop reports input and use GET
dev_csv_reports = ipf.fetch_all('tables/inventory/devices', reports=True, export='csv')
"""CSV export does not return reports, parameter has been excluded."""
print(type(dev_csv_reports))  # bytes

# If URL exceeds 4096 characters the following exception will be raised:
# raise InvalidURL(f"URL exceeds max character limit of 4096: length={len(url)}.")

Installation

pip install ipfabric

Introduction

Please take a look at API Programmability - Part 1: The Basics for instructions on creating an API token.

Most of the methods and features can be located in Examples to show how to use this package. Another great introduction to this package can be found at API Programmability - Part 2: Python

Diagrams

Diagramming in IP Fabric version v4.3 and above has been moved to it's own package.

Diagramming will move back to this project in v7.0

pip install ipfabric-diagrams

Authentication

Username/Password

Supply in client:

from ipfabric import IPFClient
ipf = IPFClient('https://demo3.ipfabric.io/', auth=('user', 'pass'))

Token

from ipfabric import IPFClient
ipf = IPFClient('https://demo3.ipfabric.io/', auth='token')

Environment

The easiest way to use this package is with a .env file. You can copy the sample and edit it with your environment variables.

cp sample.env .env

This contains the following variables which can also be set as environment variables instead of a .env file.

IPF_URL="https://demo3.ipfabric.io"
IPF_TOKEN=TOKEN
IPF_VERIFY=true

Or if using Username/Password:

IPF_URL="https://demo3.ipfabric.io"
IPF_USERNAME=USER
IPF_PASSWORD=PASS

Development

Poetry Installation

IPFabric uses Poetry to make setting up a virtual environment with all dependencies installed quick and easy.

Install poetry globally:

pip install poetry

To install a virtual environment run the following command in the root of this directory.

poetry install

To run examples, install extras:

poetry install ipfabric -E examples

Test and Build

poetry run pytest
poetry build

Prior to pushing changes run:

poetry run black ipfabric
poetry update