frr-ospf-api-stream-client

FRR OSFP API Stream Client

This repository contains a Python client for the FRR OSPF API which synchronizes a copy of the network's LSDB via LSA update streams for export into various formats:

• Fixed JSON snapshots in the format of the Bird OSFP Link Database Parser
• A continuous stream of changes, filtered from the original LSA stream to only include changes which actually influence the network graph or injected routes

This work is based in large part on the reference script included in FRR. You should start there if you are trying to develop your own custom FRR OSPF API client.

> ospf-stream-client --server <frr OSPF API server addr>
2024-12-23 04:29:32,352 WARNING: CLIENT: root Waiting for initial load to complete...
2024-12-23 04:29:32,465 WARNING: CLIENT: root LSDB loaded!
{"timestamp": 1734928168632, "entity": {"type": "router", "id": "10.69.5.52"}, "added": {"link": {"router": {"id": "10.69.6.10", "metric": 100}}}}
{"timestamp": 1734928167633, "entity": {"type": "router", "id": "10.69.6.10"}, "added": {"link": {"router": {"id": "10.69.5.52", "metric": 100}}}}
{"timestamp": 1734928166886, "entity": {"type": "router", "id": "10.69.3.85"}, "removed": {"link": {"router": {"id": "10.69.4.40", "metric": 100}}}}
{"timestamp": 1734928165886, "entity": {"type": "router", "id": "10.69.1.65"}, "added": {"link": {"router": {"id": "10.69.5.51", "metric": 100}}}}
{"timestamp": 1734928164890, "entity": {"type": "router", "id": "10.69.5.51"}, "added": {"link": {"router": {"id": "10.69.1.65", "metric": 100}}}}

Limitations

This client doesn't implement support for multi-area OSPF LSAs or IPv6 (OSPFv3) and will only work for IPv4 OSPF networks where all routers are present in the backbone network. Exact behavior for trying to use this tool on multi-area networks is undefined but expect crashes

Output Format

The DB snapshot and streamed updates output formats are detailed using JSON Schema in src/api_stream_client/db_snapshot.json and src/api_stream_client/stream_event.json respectively. Note that the stream is composed of a series of newline-delimited JSON objects, each of which is formatted according to stream_event.json. This means the stream (and any subset of its lines) is JSONL-compliant. Read more about JSONL here.

Usage

Pre-requisites: python3 available via the shell

First, install the CLI via pip:

pip install frr-ospf-api-stream-client

then invoke the tool with the CLI command:

ospf-stream-client --server <frr OSPF API server adress>

Available flags:

• --server <frr OSPF API server adress>: Specify the address that the FRR OSPF API server is running on. Defaults to localhost. Note: the FRR OSPF API sockets protocol is somewhat esoteric. It relies on bidirectional sockets being intiated with dynamic port numbers between the client and the server, so while it is technically possible to run the client on a separate host from the server, the connection will not work across most firewalls and definitely WILL NOT work through any kind of NAT
• --ws-listen <interface address>:<listen port>: If provided, we will use autobahn to listen on the specified interace and port for inbound websockets connections. These connections are very simple, they do not accept any input from clients. All clients recieve an identical copy of the event stream, one JSON event per WS message
• --events-path-prefix <folder path>: If provided, we will write the event stream to disk, paritioning into files of one hour each, within the specified prefix. You likely want some kind of cleanup job to compress/archive/delete these files to avoid filling up your disk
• --snapshots-path-prefix <folder path>: If provided, we will write LSDB snapshots to disk, once per minute grouping by year, month, day, hour, etc, within the specified prefix. You likely want some kind of cleanup job to compress/archive/delete these files to avoid filling up your disk
• --mute-stdout-stream: By default, the event stream will be written to stdout. Provide this flag to disable that behavior
• -v, --verbose: Increase the verbosity of the stderr log messages. Useful for troubleshooting upstream API issues

Dev Setup

Pre-requisites: python3 available via the shell

Setup by cloning, creating a virtual env, and installing the application

git clone https://github.com/Andrew-Dickinson/frr-ospf-api-stream-client
cd frr-ospf-api-stream-client
python3 -m venv .venv
source .venv/bin/activate
pip install -e .

then invoke the tool with the CLI command:

ospf-stream-client --server <frr OSPF API server adress>

Running the unit tests

Follow the instructions under "Dev Setup" above, to clone a local copy of this application and activate the virtual environment. Then installing the test dependencies with:

pip install -e ".[test,dev]"

Finally, invoke the test suite using pytest:

pytest test/

Building to PyPi

Follow the instructions above to clone a local copy of this application, activate the virtual environment, and run the tests.

Then, build & upload the application with

rm -rf dist/*
python -m build .
twine upload dist/*

License

Distributed under the GPL-2.0 License. See license.txt for more information.

Contributing

Contributions are what make the open source community such an amazing place to learn, inspire, and create. Any contributions you make are greatly appreciated.

If you have a suggestion that would make this better, please fork the repo and create a pull request. You can also simply open an issue with the tag "enhancement". Don't forget to give the project a star! Thanks again!

• Fork the Project
• Create your Feature Branch (git checkout -b feature/AmazingFeature)
• Commit your Changes (git commit -m 'Add some AmazingFeature')
• Push to the Branch (git push origin feature/AmazingFeature)
• Open a Pull Request

Acknowledgments

• Best-README-Template
• FRR OSPF API Python Example
• RFC 2328
• autobahn