Process Management¶

erpc.py provides three process managers for different deployment scenarios.

ERPCProcess¶

The standard subprocess manager. Runs eRPC as a local child process.

Context Manager (Recommended)¶

from erpc import ERPCProcess

with ERPCProcess(upstreams={1: ["https://eth.llamarpc.com"]}) as erpc:
    print(erpc.endpoint_url(1))
    print(f"PID: {erpc.pid}")
    print(f"Healthy: {erpc.is_healthy}")
# eRPC is automatically stopped on exit

Manual Lifecycle¶

from erpc import ERPCConfig, ERPCProcess

config = ERPCConfig(upstreams={1: ["https://eth.llamarpc.com"]})
proc = ERPCProcess(config=config)

proc.start()
proc.wait_for_health(timeout=30)

# ... use eRPC ...

proc.stop()

Shorthand Initialization¶

Pass upstreams directly without creating a config:

proc = ERPCProcess(upstreams={1: ["https://eth.llamarpc.com"]})

Custom Binary Path¶

proc = ERPCProcess(
    upstreams={1: ["https://eth.llamarpc.com"]},
    binary_path="/opt/erpc/erpc",
)

Health Checks¶

with ERPCProcess(upstreams={1: ["https://eth.llamarpc.com"]}) as erpc:
    # Boolean health check
    if erpc.is_healthy:
        print("eRPC is up")

    # Structured health via client
    client = erpc.client
    status = client.health()
    print(f"Version: {status.version}, Uptime: {status.uptime}s")

Process Signals¶

import signal

proc = ERPCProcess(upstreams={1: ["https://eth.llamarpc.com"]})
proc.start()

# Graceful shutdown (SIGTERM)
proc.stop()

# Force kill (SIGKILL) — if stop() doesn't work
proc.stop(force=True)

AsyncERPCProcess¶

For asyncio applications. Same API as ERPCProcess but with async context manager:

import asyncio
from erpc import AsyncERPCProcess

async def main():
    async with AsyncERPCProcess(upstreams={1: ["https://eth.llamarpc.com"]}) as erpc:
        print(erpc.endpoint_url(1))
        print(f"Healthy: {erpc.is_healthy}")

asyncio.run(main())

DockerERPCProcess¶

Runs eRPC in a Docker container. No local binary needed — uses the docker CLI.

Basic Usage¶

from erpc import ERPCConfig, DockerERPCProcess

config = ERPCConfig(upstreams={1: ["https://eth.llamarpc.com"]})

with DockerERPCProcess(config=config, name="my-erpc") as erpc:
    print(erpc.endpoint_url(1))
    print(erpc.logs(tail=20))

Custom Image¶

proc = DockerERPCProcess(
    config=config,
    name="erpc-prod",
    image="ghcr.io/erpc/erpc:0.0.62",
)

Container Logs¶

with DockerERPCProcess(config=config) as erpc:
    # Get recent logs
    logs = erpc.logs(tail=50)
    print(logs)

Lifecycle Summary¶

All process managers follow the same pattern:

Method	Description
`start()`	Start the eRPC process/container
`stop()`	Gracefully stop eRPC
`wait_for_health()`	Block until healthy or timeout
`is_healthy`	Property — quick boolean health check
`endpoint_url(chain_id)`	Get the RPC URL for a chain
`client`	Get an `ERPCClient` bound to this instance

Context Manager

Using with (or async with) is strongly recommended. It ensures cleanup even if exceptions occur.