opcua-sim – simulated OPC UA devices in your browser

pump (drive) frequency converter –

– – –

state– set point– motor current–

powering up…

tank_pressure (transmitter) 4–20 mA, NE 43 –

– – –

loop current– status–

powering up…

oven (controller) PI loop –

– – –

set point– output–

powering up…

meter (energy meter) three phases –

L1– L2– power– energy–

powering up…

plc (S7-1500 layout, simulated) ns=3;s="DB_Line"."Speed"

Run– Speed– Parts– Fault–

powering up…

reader (BCL 300 layout, simulated) ns=4;i=6205

status– last code– good reads– bad reads–

powering up…

no session starting…

Alarms retained on this server:

none

Try these in order:

Break something:

What you just used

your SCADA, HMI, gateway or Python code opc.tcp://…:4840 → opcua-sim instead of the plant

What runs here and what does not. The devices, their behaviour, faults and alarm states on this page are the real opcua-sim code. The OPC UA wire protocol is not: a browser has no TCP sockets, so your commands reach the devices directly and the tx lines name the service, not the bytes. This page answers by the same rules, but to test a client you install opcua-sim. There it is a real server on opc.tcp://, built on asyncua, checked with asyncua and node-opcua clients.

Health. The badge on each device is its DeviceHealth node, with the NAMUR NE 107 states: normal, failure, check function, off spec, maintenance required.

Product layouts. The S7-1500 and BCL 300 servers copy how those products present their nodes, as far as public documents describe it. They are not complete copies and were not compared with real hardware. Each profile names its sources and what is assumed.

The same thing against a real OPC UA server

pip install opcua-sim
opcua-sim serve --bench plant.yaml      # the plant from this page, on opc.tcp://127.0.0.1:4840

import asyncio
from asyncua import Client, ua

async def main():
    async with Client("opc.tcp://127.0.0.1:4840") as client:
        idx = await client.get_namespace_index("urn:steitec:opcua-sim")
        methods = client.get_node(ua.NodeId("pump.MethodSet", idx))
        await methods.call_method(ua.NodeId("pump.Start", idx))
        await asyncio.sleep(2)
        print(await client.get_node(ua.NodeId("pump.Speed", idx)).read_value())   # about 20.0

asyncio.run(main())

And in a test, with a new server for each test:

def test_notices_a_device_failure(opcua_bench):
    opcua_bench.inject("drive", "bad_status", variable="Speed", status="BadDeviceFailure")
    with pytest.raises(ua.uaerrors.BadDeviceFailure):
        read_speed(opcua_bench.url)       # your code

What else it does

Real alarms: level and discrete alarms as OPC UA events, Acknowledge with the right EventId, ConditionRefresh after a reconnect.
Faults: any status code on any variable, frozen values, old timestamps, drift, refused writes and calls, late answers, dropped and refused connections, server state and service level.
Security: None, Sign and SignAndEncrypt endpoints, user names, read-only users.
Your own devices from a YAML file. opcua-sim map prints the address space.
Profiles: pressure transmitter, temperature controller, drive, energy meter, generic sensor, and layouts after Siemens S7-1500, Beckhoff TwinCAT 3, Phoenix PLCnext, CODESYS, Omron NJ/NX and Leuze BCL 300.

Source on GitHub PyPI Questions or ideas? Open an issue