Persisting Models

Evolve data service

danger

The API for persisting data to a remote server is currently in alpha and very likely to experience breaking changes in the future. Please provide any feedback about this API to Zepben.

The SDK provides a client to send models to a remote data server via gRPC. The service and proto definitions for this API can be found here. An implementation of the server for the services is provided with the Evolve platform data services.

Here is an example of how to connect to and send a model to the evolve data server:

from zepben.evolve import connect_async, connect, ProducerClient, NetworkService, CustomerService, DiagramService

async with connect_async(host="localhost", rpc_port=50051) as channel:
    client = ProducerClient(channel)

    network_service = NetworkService()
    customer_service = CustomerService()
    diagram_service = DiagramService()
    populate_network_objects(network_service)

    await client.send([network_service, customer_service, diagram_service])

# Or using the synchronous API:
with connect(host="localhost", rpc_port=50051) as channel:
    client = ProducerClient(channel)

    network_service = NetworkService()
    customer_service = CustomerService()
    diagram_service = DiagramService()
    populate_network_objects(network_service)

    client.send([network_service, customer_service, diagram_service])

The ProducerClient is an abstraction layer that allows sending all services at once (useful for a server that exposes more than one service), however typically you'll want to use a dedicated service for the data you are working with. For these cases there are the following classes:

• NetworkProducerClient
• CustomerProducerClient
• DiagramProducerClient

These classes expose a similar API to ProducerClient, however accepting a single service of the named type rather than a list. For example:

from zepben.evolve import connect_async, NetworkService, CustomerService, DiagramService, NetworkProducerClient, CustomerProducerClient, DiagramProducerClient

async with connect_async(host="localhost", rpc_port=50051) as channel:
    network_client = NetworkProducerClient(channel)

    network_service = NetworkService()
    populate_network_objects(network_service)

    await client.send([network_service, customer_service, diagram_service])

async with connect_async(host="localhost", rpc_port=50052) as channel:
    customer_client = CustomerProducerClient(channel)

    customer_service = CustomerService()
    populate_customer_objects(customer_service)

    await client.send(customer_service)

async with connect_async(host="localhost", rpc_port=50052) as channel:
    diagram_client = DiagramProducerClient(channel)

    diagram_service = DiagramService()
    populate_diagram_objects(diagram_service)

    await client.send(diagram_service)