This package is a python wrapper for leveraging the API for Aruba Orchestrator and Edge Connect SDWAN systems.
API's are documented via Swagger directly on Orchestrator and Edge Connect web interfaces under "Support > Rest API"
Many, but not all API functions have been implemented yet. Development is underway to continue to add further functions.
THERE IS NOW PRELIMINARY SUPPORT FOR API CHANGES INTRODUCED IN ORCHESTRATOR 9.3+
Note: Requires Python 3.7+
Once Python 3.7+ is installed on the system, it's recommended to use a virtual environment to install the package to.
In the directory you'd like to write your project/script, setup a python virtual environment with the desired python version and activate it. This is important if you have other versions of python installed on your system.
:~$ python3.9 -m venv my_new_project
:~$ source my_new_project/bin/activate
(my_new_project) :~$ python --version
Python 3.9.13
Now you are ready to install the package and run your python code.
Note: Going forward, these commands assume you're within a Python 3.7+ venv, or Python 3.7+ is the exclusive Python version installed in regard to referencing the use of
pip
. If that is not the case, you can specifically appendpython3.x -m
ahead of thepip install ...
$ pip install pyedgeconnect
...
$ pip list
Package Version
----------------------------- --------------------------------
... ...
pyedgeconnect x.y.z
... ...
To install the most recent version of pyedgeconnect, open an interactive shell and run:
$ pip install git+https://github.com/aruba/pyedgeconnect
...
$ pip list
Package Version
----------------------------- --------------------------------
... ...
pyedgeconnect x.y.z
... ...
To install a specific branch use the @branch syntax
$ pip install git+https://github.com/aruba/pyedgeconnect@<branch_name>
...
# Install the Development branch
$ pip install git+https://github.com/aruba/pyedgeconnect@development
...
For editing the code and general testing you can specify the [dev]
extras which will include ["black", "flake8", "flake8-rst-docstrings",
"isort", "sphinx", "sphinx_rtd_theme"]
To install from the remote repo with the [dev]
extras option use the
following syntax:
$ pip install pyedgeconnect[dev]
or
$ pip install git+https://github.com/aruba/pyedgeconnect#egg=pyedgeconnect[dev]
Docs are viewable on Read the Docs
To build the documentation locally, clone the repository, install with [dev]
option
to include sphinx and related packages, then in the docs directory run make html
git clone https://github.com/aruba/pyedgeconnect.git
cd edgeconnect-python
pip install .[dev]
cd docs
make html
Import the Orchestrator class to your script.
from pyedgeconnect import Orchestrator
To initialize an Orchestrator you must pass the url of the Orchestrator (IP or FQDN).
Note: If you're connecting to an Orchestrator without a valid certificate you'll want to set the
verify_ssl
paramter toFalse
when instantiating Orchestrator to ignore certificate warnings/errors.
orch_url = '10.1.1.100'
orch_url = 'orchestrator.example.com'
orch = Orchestrator(orch_url, verify_ssl=False)
Now you can call the login function to connect to Orchestrator with a username and password:
orch_user = 'admin'
orch_pw = 'change_me'
orch.login(orch_user, orch_pw)
orch.logout()
Another option is to pass an API Key on init to make authenticated calls without having to call login/logout functions
orch_url = 'orchestrator.example.com'
orch = Orchestrator(orch_url, api_key='xxx')
from pyedgeconnect import EdgeConnect
To initialize an Edge Connect you must pass the url of the Edge Connect (IP or FQDN).
Note: If you're connecting to an Edge Connect without a valid certificate you'll want to set the
verify_ssl
paramter toFalse
when instantiating EdgeConnect to ignore certificate warnings/errors.
ecos_url = '10.2.30.50'
ecos_url = 'edgeconnect.example.com'
ec = EdgeConnect(ecos_url, verify_ssl=False)
Now you can call the login function to connect to Edge Connect with a username and password:
ecos_user = 'admin'
ecos_pw = 'admin'
ec.login(ecos_user, ecos_pw)
ec.logout()
By default, Orchestrator and EdgeConnect classes will not log calls and/or errors to file or console. When instantiating Orchestrator or EdgeConnect classes follow the below settings to enable logging options:
-
Logging to a local file: set the
log_file
parameter toTrue
. This will create ./logging/sp_orch.log or ./logging/sp_ecos.log relative to where python is launched for calls that are performed. -
Logging to the console: set the
log_console
parameter toTrue
By default, successful API calls (e.g. returning HTTP 200/204 etc.) will
not log response text to avoid logging sensitive data. To include
response text in log messages, set the log_success
parameter to
True
.
Warning: If
log_file
andlog_success
are set toTrue
response text from successful API calls will be logged to the local log file. Some responses can include sensitive data that you may not wish to retain in the log files.
orch_url = 'orchestrator.example.com'
orch = Orchestrator(orch_url, log_file=True, log_console=True)
# or
ecos_url = '10.2.30.50'
ec = EdgeConnect(ecos_url, log_success=True)
In the Examples directory you can find scripts leveraging the Orchestrator class demonstrating some uses
- create_user.py
- creates a new read-only user on Orchestrator and returns the configured details
- print_appliance_info.py
- retrieves all appliances, retrieves detailed attributes of the appliances, and prints details in a table format
- run_packet_capture.py
- runs a tcpdump packet capture on a specified appliance, once completed, uploads to Orchestrator for user retrieval
- preconfig.py
- uses a CSV file as source data to generate Edge Connect YAML preconfig from a Jinja template
This package is still very new. This is made explicit by the "Alpha" trove classifier, as well as by the "a" in the version number. Until the package becomes stable, you should expect some formatting and/or syntax to change in the future.
MIT
Adding more modules and API functions are prioritized as needed for use. There is not currently support for reviewing external PR's as maintenance is best effort by the authors.
Open an issue to track any/all suggestions/fixes/additions. Please don't file an issue to ask a question.
As this code is in early stages there are larger changes that may be discussed in regards to overall structure, error handling, logging, etc. Suggestions for these topics can be raised via issue or contacting the authors.
See contribution details at Contributing
Release notes are located in docs/source/release-notes
directory here
Authored by Zach Camara, email at zachary.camara@hpe.com