/aiognmi

Async, efficient and lightweight gNMI client written in Python

Primary LanguagePythonMIT LicenseMIT

Supported Versions PyPI version Ruff License: MIT CI

aiogNMI

About

This Python library provides an efficient and lightweight gNMI client implementation that leverages asynchronous approach.

Supported RPCs:

  • Capabilities
  • Get
  • Set
  • Subscribe (under development)

Tested on:

  • Arista EOS
  • Nokia SR OS

Repository contains protobuf files from gNMI repo and based on gNMI release v0.10.0. Early gNMI version should work too, I've tested with 0.7.0 and it works well.

NOTE: At this moment supporting of the secure connections (with encryption or certificate) is in alpha version. You can use them, but I don't guarantee stable work.

Install

Install with pip:

pip install aiognmi

Examples

Capabilities RPC

import asyncio

from aiognmi import AsyncgNMIClient


async def main():
    async with AsyncgNMIClient(host="test-1", port=6030, username="admin", password="admin", insecure=True) as client:
        resp = await client.get_capabilities()

    print(resp.result)


if __name__ == "__main__":
    asyncio.run(main())

Get RPC

import asyncio

from aiognmi import AsyncgNMIClient


async def main():
    async with AsyncgNMIClient(host="test-1", port=6030, username="admin", password="admin", insecure=True) as client:
        resp = await client.get(
            paths=[
                "/interfaces/interface[name=Management0]",
            ]
        )

    print(resp.result)


if __name__ == "__main__":
    asyncio.run(main())

Set RPC

import asyncio

from aiognmi import AsyncgNMIClient


async def main():
    async with AsyncgNMIClient(host="test-1", port=6030, username="admin", password="admin", insecure=True) as client:
        resp = await client.set(
            update=[
                {"path": "/interfaces/interface[name=Management0]/config", "data": {"description": "gnmi update test"}}
            ]
        )

    print(resp.result)


if __name__ == "__main__":
    asyncio.run(main())

Credits

My work is inspired by these people:

  1. Anton Karneliuk and his pyGNMI library
  2. Carl Montanari and his scrapli library