Skip to main content

NuroPb - A Distributed Event Driven Service Mesh

Project description

NuroPb

The nervous system and plumbing for an Asynchronous, Distributed, Event Driven, Service Mesh

codecov Code style: black CodeFactor License: Apache 2.0

Do you have a Python class that you want to make available as a service to consumers or other services, or

  • You potentially want to scale a service horizontally many times over, likely at an unknown scale.
  • Your service may need to communicate to other services
  • There are event driven processes and flows across your service landscape
  • You have websocket endpoints that need to integrate seamlessly across a service mesh and event driven architecture
  • A growing army of Ml-Ops and Datascience engineers are joining, with the need to to integrate their efforts across many systems.
  • Needing a service proxy that bridges cloud VPNs and on-premise networks

If any of these are of interest to you, then NuroPb is worth considering. NuroPb It falls into the domain of
tools and frameworks for the abstraction of integration plumbing, enabling software engineers to focus on the problems they're hired to solve.

Where does the name originate from?

NuroPb is a contraction of the term nervous system and the scientific symbol for Lead. Lead associated with plumbing. So NuroPb is a system's nervous system and plumbing.

Pattern and Approach

NuroPb is a pattern and approach that supports event driven and service mesh engineering requirements. The early roots evolved in the 2000's, and during the 2010's the pattern was used to drive hedge funds, startups, banks and crypto and blockchain ventures. It's core development is in Python, and the pattern has been used alongside Java, JavasScript and GoLang, Unfortunately those efforts are copyrighted. Any platform with support for RabbitMQ should be able to implement the pattern and exist on the same mesh as Python Services.

RabbitMQ is the underlying message broker for the NuroPb service mesh library. Various message brokers and broker-less tools and approaches have been tried with the nuropb pattern. Some of these are Kafka, MQSeries and ZeroMQ. RabbitMQ's AMPQ routing capabilities, low maintenance and robustness have proved the test of time. With RabbitMQ's release of the streams feature, and offering message/logs streaming capabilities, there are new and interesting opportunities all on a single platform.

Why not Kafka? Kafka is a great tool, but it's not a message broker. It's a distributed log stream and probably one of the best available. There are many use cases where NuroPb + RabbitMQ would integrate very nicely alongside Kafka. Especially inter-process rpc and orderless event driven flows that are orchestrated by NuroPb and ordered log/event streaming over Kafka. Kafka is also been used as to ingest all nuropb traffic for auditing and tracing.

Getting started

The best way to get started is to look at the examples and the examples/README.md file.

As a super quick reference this si what you need to get started:

  • Python >= 3.10
  • RabbitMQ >= 3.8.0 + Management Plugin
  • pip install nuropb

This code block is an example of a client and server running in the same python file.

import logging
from typing import Any, Dict
from uuid import uuid4
import asyncio

from nuropb.contexts.context_manager import NuropbContextManager
from nuropb.contexts.context_manager_decorator import nuropb_context
from nuropb.contexts.describe import publish_to_mesh
from nuropb.rmq_api import RMQAPI

logger = logging.getLogger("nuropb-all-in-one")


def get_claims_from_token(bearer_token: str) -> Dict[str, Any] | None:
    """ This is a stub for the required implentation of validating and decoding the bearer token
    """
    _ = bearer_token
    return {
        "sub": "test_user",
        "user_id": "test_user",
        "scope": "openid, profile",
        "roles": "user, admin",
    }


class QuickExampleService:
    _service_name = "quick-example"
    _instance_id = uuid4().hex

    @nuropb_context
    @publish_to_mesh(authorise_func=get_claims_from_token)
    def test_requires_user_claims(self, ctx, **kwargs: Any) -> str:
        logger.info("test_requires_user_claims called")
        assert isinstance(ctx, NuropbContextManager)
        return f"hello {ctx.user_claims['user_id']}"

    def test_method(self, param1, param2: Dict[str, Any]) -> Dict[str, Any]:
        logger.info("test_method called")
        _ = self
        return {
            "param1": param1,
            "param2": param2,
            "reply": "response from test_method",
        }


async def main():
    logging.info("All in one example done")
    amqp_url = "amqp://guest:guest@localhost:5672/nuropb-example"
    service_instance = QuickExampleService()
    transport_settings = {
        "rpc_bindings": [service_instance._service_name],
    }
    service_api = RMQAPI(
        service_instance=service_instance,
        service_name=service_instance._service_name,
        instance_id=service_instance._instance_id,
        amqp_url=amqp_url,
        transport_settings=transport_settings,
    )
    await service_api.connect()
    logger.info("Service Ready")

    client_api = RMQAPI(
        amqp_url=amqp_url,
    )
    await client_api.connect()
    logger.info("Client connected")

    context = {
        "Authorization": "Bearer 1234567890",
    }
    response = await client_api.request(
        service="quick-example",
        method="test_requires_user_claims",
        params={},
        context=context,
    )
    logger.info(f"Response: {response}")

    response = await client_api.request(
        service="quick-example",
        method="test_method",
        params={
            "param1": "value1",
            "param2": {
                "param2a": "value2a",
            }
        },
        context={},
    )
    logger.info(f"Response: {response}")

    await client_api.disconnect()
    await service_api.disconnect()

    logging.info("All in one example done")


if __name__ == "__main__":
    log_format = (
        "%(levelname).1s %(asctime)s %(name) -25s %(funcName) "
        "-35s %(lineno) -5d: %(message)s"
    )
    logging.basicConfig(level=logging.INFO, format=log_format)
    logging.getLogger("pika").setLevel(logging.WARNING)
    logging.getLogger("etcd3").setLevel(logging.WARNING)
    logging.getLogger("urllib3").setLevel(logging.WARNING)
    asyncio.run(main())

Project details


Download files

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

Source Distribution

nuropb-0.1.6.tar.gz (70.9 kB view details)

Uploaded Source

Built Distribution

nuropb-0.1.6-py3-none-any.whl (58.0 kB view details)

Uploaded Python 3

File details

Details for the file nuropb-0.1.6.tar.gz.

File metadata

  • Download URL: nuropb-0.1.6.tar.gz
  • Upload date:
  • Size: 70.9 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? No
  • Uploaded via: poetry/1.5.1 CPython/3.11.4 Darwin/22.6.0

File hashes

Hashes for nuropb-0.1.6.tar.gz
Algorithm Hash digest
SHA256 e2f81854c97ceb38f58c2a380ba84e38bff331de7800bd6a97e6353c80994163
MD5 13198ab7c5da04bc52b2c4bf361a9b2e
BLAKE2b-256 cec405654d40a3c83472f5a1b88b326011282814761ec5465230fe5daf7f908c

See more details on using hashes here.

File details

Details for the file nuropb-0.1.6-py3-none-any.whl.

File metadata

  • Download URL: nuropb-0.1.6-py3-none-any.whl
  • Upload date:
  • Size: 58.0 kB
  • Tags: Python 3
  • Uploaded using Trusted Publishing? No
  • Uploaded via: poetry/1.5.1 CPython/3.11.4 Darwin/22.6.0

File hashes

Hashes for nuropb-0.1.6-py3-none-any.whl
Algorithm Hash digest
SHA256 51c56991a32ce06a4ddf320ecc5594abef518a57283d4436ef5c8877b35d2a9e
MD5 a989132b82cd028b275c92d97ba571a9
BLAKE2b-256 02c24b9c1f91f3b643c510783a0acd2c875c2bff077eb643c11c9cb6f84b280f

See more details on using hashes here.

Supported by

AWS AWS Cloud computing and Security Sponsor Datadog Datadog Monitoring Fastly Fastly CDN Google Google Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Sentry Sentry Error logging StatusPage StatusPage Status page