Skip to main content

Publish Docs Publish stable release to PyPi.org Publish to test.pypi

Preamble

I wrote this package (dcnodatg) specifically to assist in fast/low-effort modeling of a production EVPN Layer-3 Leaf & Spine physical network of Arista switches in a pre-existing GNS3 server, using Arista cEOS docker images. (Also, to teach myself how to write/publish Python packages, but really for modeling Arista switches.)

What it does

  • Grabs startup configuration, version info, and lldp neighbor information from a list of Arista switches
  • Uses Arista eAPI (credentials must be provided as arguments) to retrieve all data
  • Sanitizes the switch configs for use in a cEOS environment
    • Removes all AAA and username configuration
    • Reformats interface names to match the cEOS interface naming convention Ethernet_n , not Ethernt_n/1
    • Comments out incompatible commands ("queue..." not supported on cEOS)
    • Configures the system mac-address of the production switch
      • Increase verisimilitude with prod device that is being modeled
      • Avoids mLAG incompatibility with cEOS
        • Docker container default mac address has U/L bit set to L instead of U, which prevents MLAG from working
  • Builds a table of interconnections between the switches
    • Inferred from the "show lldp neighbor" and "show lldp local" output
  • Creates a GNS3 project
    • Instantiates a cEOS container node in the project for each switch in the input list
    • Modeled devices mirror the following properties of the switches they are modeling:
      • cEOS version (a pre-existing GNS3 docker template using the matching cEOS version must be present)
      • Ethernet interface count
      • system-mac-address
      • Startup configuration
        • "startup-config" is pushed from the dcnodatg package directly to the containerd service on the GNS3 server
          • Avoiding the need for dcnodatg to run on the gns3 server
    • Creates the connections between the GNS3/cEOS nodes, mirroring all inter-switch connections discovered in LLDP tables

What you'll need

Python

  • The dcnodatg project was written using Python 3.12.
    • I haven't tested it with any other versions.
  • The host running the dcnodatg packages will need to have Python and the packages listed in the dependencies section of pyproject.toml installed
  • Once Python is installed, use pip to install either the test or "stable" version of dcnodatg (which will install its dependencies as well):

GNS3 server

  • The dcnodatg module was written against version 2.2.18 of GNS3 server.
  • The GNS3 server must be pre-configured with cEOS docker templates
    • dcnodatg will compare the EOS version string on the switches you tell it to process to the tags of the docker-images on the GNS3 server
      • The docker images need to be named as "ceos:n.n..." for the matching to work
  • The containerd service must be listening on TCP port 2375 of the GNS3 server

Arista Switches

All switches that you will be modeling will need to have:

  • EAPI services accessible from the host running the dcnodatg module
    • And you will need to provide auth. credentials with sufficient privileges to invoke the following methods:
      • node.enable(("show version", "show lldp neighbors", "show lldp local-info"), format="json")
      • node.startup_config.splitlines()

Instructions

Prep

  • Have Python and the dcnodatg package installed on the host that will run dcnodatg
  • Have your login credentials for your production switches handy
  • Make sure that your production switches can receive eAPI connections from your GNS3 server
  • Optionally, create a file named "input-switch-list"
      • Populate 'input-switch-list' with the names of the switches that you want to model in gns3
      • One switch name per line (no quotes or commas)

Parameter/argument list

dcnodatg uses the following arguments (passed as keyword pairs):

  • filename or switchlist
    • No default value
    • If both arguments are provided, dcnodatg will exit.
    • If no argument is provided, dcnodatg will try use the input function to prompt for switch names
    • "filename" is (path and) name of the file containing the list of switches to process
      • One switch-name (FQDN or resolvable short-name) per line in the file
      • E.G.: ./switch-list.txt
    • "switchlist" is a python list of switch-names
      • E.g.: ["name1", "name2", "nameN"]
  • servername
    • No default value
    • The name (FQDN or resolvable short-name) of the GNS3 server
    • If not provided, dcnodatg will try to use the input function to prompt for a value
    • E.g.: gns3server.whathwere.there
  • username
    • No default value
    • The username dcnodatg will provide to the switches when authenticating the eapi connections
    • If not provided, dcnodatg will try to use the input function to prompt for a value
  • passwd
    • No default value
    • The password dcnodatg will provide to the switches when authenticating the eapi connections
    • If not provided, dcnodatg will try to use the input function to prompt for a value
  • prjname
    • No default value
    • The name to assign to the new project that dcnodatg will create on the GNS3 server

Execution

As a Python script

Installing dcnodatg via pip will save you the effor of installing the additional dependencies list in pyproject.toml, but you can also just grab the contents of the dcnodatg folder directly from the git repository and store them on the host you'll run them from.

You'll also need to move the "dcnod-cli.py" file up one level in the directory structure from the dcnodatg folder after copying the entire folder to your host. This is to work around "goofiness" with regards to how Python treats namespaces when accessing Python code as a "script" vs accessing it "as a module."

To actually run the utility, you'll enter the following command:

python [path-to]dcnod-cli.py'
To run interactively

Enter:

python [path-to]dcnod-cli.py'

As dcnodatg executes, you will be prompted to respond with values for all of the parameters/arguments. No quotes or delimiters should be required as you enter the values.

  • The FQDNs of the switches you want to process
    • Type a switch-name and press Enter
    • Repeat until you've entered all the switches you want to model
    • Then press Enter again
  • The name of the GNS3 project to create
    • Type a project name (adhere to whatever GNS3's project-naming semantics) and press enter
  • The EOS username to use when querying the switches
    • Type the name and press enter
  • The EOS password to use when querying the switches
    • The getpass function is used, obscuring the password on-screen as you type it
    • The password itself isn't written to any file
    • Type the password and press Enter
  • The FQDN of the GNS3 server you'll be using
To run non-interactively

Enter:

python [path-to]dcnod-cli.py [arguments]

The arguments are keyword/value pairs, in the form of:

keyword='value'

The arguments can be entered in any order, separate by a space. Examples of each argument follow:

username= 'mynameismud'
passwd= 'mypasswordisalsomud'
servername= 'gns3server.menckend.com'
switchlist= 'sw1.menckend.com sw2 sw3 sw4.menckend.com'
filename= './switchlist.txt'
prjname= 'dcnodatg-project-dujour'

Remember that the switchlist and filename arguments are mutually exclusive, if you pass both, dcnodatg will exit.

An example of a fully-argumented invocation would be:

python ./dcnodatg.py username='fakeid' passwd='b@dp@ssw0rd' servername='gn3server.com' prjname='giveitanme' switchlist='switch1 switch2 switch3'
As a Python module

Install dcnodatg with pip as described above and include an import statement ('import dcnodatg') in your python module. E.g.

from dcnodatg import dcnodatg

sn='gns3server.bibbity.bobbity.boo'
un='myuserid'
pw='weakpassword'
prjn='new-gns3-project-today'
sl=['switch1.internal', 'switch15.internal', 'switch1.menckend.com']

dcnodatg.p_to_v(username=sn, passwd=pw, servername=sn, switchlist=sl, prjname=prjn)

[!IMPORTANT]
The 'switchlist' parameter, when dcnodatg is being accessed as a module is a dict structure, and the formatting in the example above is mandatory when specifying the switchlist data as a kwarg.

Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

dcnodatg-1.0.2.5.tar.gz (26.6 kB view details)

Uploaded Source

Built Distribution

If you're not sure about the file name format, learn more about wheel file names.

dcnodatg-1.0.2.5-py3-none-any.whl (28.0 kB view details)

Uploaded Python 3

File details

Details for the file dcnodatg-1.0.2.5.tar.gz.

File metadata

  • Download URL: dcnodatg-1.0.2.5.tar.gz
  • Upload date:
  • Size: 26.6 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? Yes
  • Uploaded via: twine/5.1.1 CPython/3.12.7

File hashes

Hashes for dcnodatg-1.0.2.5.tar.gz
Algorithm Hash digest
SHA256 10bc9fbc5e665ceb081b0419aa5d70a07662a32fe9bd4a4fcad8ea31fab1753f
MD5 521afd5e46479b12c5a52ae7bdb8b85e
BLAKE2b-256 17dbcec5a4171e74c64356e8a37e581d9bbfa74e36bb33cd49be486140e37ac5

See more details on using hashes here.

File details

Details for the file dcnodatg-1.0.2.5-py3-none-any.whl.

File metadata

  • Download URL: dcnodatg-1.0.2.5-py3-none-any.whl
  • Upload date:
  • Size: 28.0 kB
  • Tags: Python 3
  • Uploaded using Trusted Publishing? Yes
  • Uploaded via: twine/5.1.1 CPython/3.12.7

File hashes

Hashes for dcnodatg-1.0.2.5-py3-none-any.whl
Algorithm Hash digest
SHA256 49bf6c112ea84601988543d0076a306e8b85e2e40aa503ae536b707367dfb420
MD5 8e5c288973463ff64ab5b49fb50b93c2
BLAKE2b-256 06e7e7005f23b336297fd49333ead14b9cd2cb4bbb31283c712848983c7bd15c

See more details on using hashes here.

Supported by

AWS Cloud computing and Security Sponsor Datadog Monitoring Depot Continuous Integration Fastly CDN Google Download Analytics Sentry Error logging StatusPage Status page