Many engineers with personally-owned home labs are looking for a cost effective way to perform automated validation of their environment, especially if the equipment is not commonly used. This project is especially designed for old hardware that does not support modern APIs or integration with monitoring systems.
Contact information:
Email: njrusmc@gmail.com
Twitter: @nickrusso42518
Testing was conducted on the following platforms and versions:
- Cisco Catalyst 3750X, version 15.2(4)E10
- VMware ESXI, version 6.0 Update 2 Build 3620759
- Synology RS-814, version 6.2.2
The code is likely to work on these products running different versions without any problems as only SSH/CLI is used; no modern API interactions.
First, install the required Python packages. I'd recommend using Python 3.10
or newer, but Python 3.7 will likely also work: pip install -r requirements.txt
You can also use pip install -r requirements.freeze.txt to exactly
match my versions, but you may miss out on package updates.
This is a "some assembly required" project. Next, create config.json and
populate it with your connectivity parameters. An example is included in the
samples/ directory. This project assumes that all of your lab devices share
the same username and password for consistency. If that's not true, you'll
have to find the auth_username and auth_password for each host individually,
while removing it from the login dictionary.
Last, you should open the validate.py script and carefully review the
different test cases. Be sure to change the version, serial number, VLAN,
and other strings that are specific to your environment. I strongly encourage
you to add, delete, and modify various tests. The existing script is very
specific to my environment and the aspects important to me. This will differ
for each user.
If you find the test-and-assert process too difficult at first, you can simply
send commands to the devices as a first step. Each supported platform has a
nested _send_cmd_* coroutine that simplifies the communication process,
automatically parsing the output (if possible), and writing the results to
a host-specific log file in human-readable form. You can quickly and reliably
collect data from your device, review the log files, then determine which
fields are worth evaluating. Log files from my personal environment are
included in the samples/ directory.