Edit

Share via

Discover OPC UA data sources using Azure IoT Akri Preview

  • Article

Important

Azure IoT Operations Preview – enabled by Azure Arc is currently in PREVIEW. You shouldn't use this preview software in production environments.

See the Supplemental Terms of Use for Microsoft Azure Previews for legal terms that apply to Azure features that are in beta, preview, or otherwise not yet released into general availability.

In this article, you learn how to discover OPC UA data sources. After you deploy Azure IoT Operations, you configure Azure IoT Akri Preview to discover OPC UA data sources at the edge. Azure IoT Akri creates custom resources into the Azure IoT Operations namespace on your cluster to represent the discovered devices. The capability to discover OPC UA data sources simplifies the process of manually configuring them from the cloud and onboarding them to your cluster. Currently, Azure IoT Akri can't detect and create assets that can be ingested into the Azure Device Registry Preview. For more information on supported features, see Azure IoT Akri overview.

Azure IoT Akri enables you to detect and create Assets in the address space of an OPC UA server. The OPC UA asset detection generates AssetType and Asset Kubernetes custom resources (CRs) for OPC UA Device Integration (DI) specification compliant Assets.

Prerequisites

  • Azure IoT Operations Preview installed. The installation includes Azure IoT Akri. For more information, see Quickstart: Deploy Azure IoT Operations Preview to an Arc-enabled Kubernetes cluster.

  • Ensure that Azure IoT Akri agent pod is properly configured by running the following code:

    kubectl get pods -n azure-iot-operations

    You should see the agent and discovery handler pod running.

    NAME                                             READY   STATUS    RESTARTS   AGE
aio-akri-agent-daemonset-hwpc7                   1/1     Running   0          17m
akri-opcua-asset-discovery-daemonset-dwn2q       1/1     Running   0          8m28s

Configure the OPC UA discovery handler

To configure the custom OPC UA discovery handler with asset detection, first you create a YAML configuration file using the values described in this section. Before you create the file, note the following configuration details:

  • The specified server contains a sample address model that uses the Robotics companion specification, which is based on the DI specification. A model that uses these specifications is required for asset detection. The Robot contains five assets with observable variables and a DeviceHealth node that is automatically detected for monitoring.

  • You can specify other servers by providing the endpointUrl and ensuring that a security None profile is enabled.

  • To enable Azure IoT Akri to discover the servers, confirm that you specified the correct discovery endpoint URL during installation.

  • Discovery URLs appear as opc.tcp://<FQDN>:50000/. To find the FQDNs of your OPC PLC servers, navigate to your deployments in the Azure portal. For each server, copy and paste the FQDN value into your discovery URLs. The following example demonstrates discovery of two OPC PLC servers. You can add the asset parameters for each OPC PLC server. If you only have one OPC PLC server, delete one of the assets.

    Name Mandatory Datatype Default Comment
    EndpointUrl true String null The OPC UA endpoint URL to use for asset discovery
    AutoAcceptUntrustedCertificates true ¹ Boolean false Whether the client auto accepts untrusted certificates. A certificate can only be auto-accepted as trusted if no non-suppressible errors occurred during chain validation. For example, a certificate with incomplete chain is not accepted.
    UseSecurity true ¹ Boolean true Whether the client should use a secure connection
    UserName false String null The username for user authentication. ²
    Password false String null The user password for user authentication. ²

    ¹ The current version of the discovery handler only supports no security UseSecurity=false and requires autoAcceptUntrustedCertificates=true.
    ² Temporary implementation until Azure IoT Akri can pass K8S secrets.

  1. To create the YAML configuration file, copy and paste the following content into a new file, and save it as opcua-configuration.yaml:

    If you're using the simulated PLC server that was deployed with the Azure IoT Operations Quickstart, you don't need to change the endpointUrl. If you have your own OPC UA servers running or are using the simulated PLC servers deployed on Azure, add in your endpoint URL accordingly.

    apiVersion: akri.sh/v0
kind: Configuration
metadata:
  name: aio-akri-opcua-asset
spec:
  discoveryHandler: 
    name: opcua-asset
    discoveryDetails: "opcuaDiscoveryMethod:\n  - asset:\n      endpointUrl: \"	opc.tcp://opcplc-000000:50000\"\n      useSecurity: false\n      autoAcceptUntrustedCertificates: true\n"
  brokerProperties: {}
  capacity: 1

  2. Apply the YAML to configure Azure Iot Akri to discover the assets:

    kubectl apply -f opcua-configuration.yaml -n azure-iot-operations

  3. To confirm that the asset discovery container is configured and started, check the pod logs with the following command:

    kubectl logs <insert aio-akri-opcua-asset-discovery pod name> -n azure-iot-operations

    A log from the aio-akri-opcua-asset-discovery pod indicates after a few seconds that the discovery handler registered itself with Azure IoT Akri:

    2023-06-07 10:45:27.395 +00:00 info: OpcUaAssetDetection.Akri.Program[0]      Akri OPC UA Asset Detection (0.2.0-alpha.203+Branch.main.Sha.cd4045345ad0d148cca4098b68fc7da5b307ce13) is starting with the process id: 1
2023-06-07 10:45:27.695 +00:00 info: OpcUaAssetDetection.Akri.Program[0]      Got IP address of the pod from POD_IP environment variable.
2023-06-07 10:45:28.695 +00:00 info: OpcUaAssetDetection.Akri.Program[0]      Registered with Akri system with Name opcua-asset for http://10.1.0.92:80 with type: Network as shared: True
2023-06-07 10:45:28.696 +00:00 info: OpcUaAssetDetection.Akri.Program[0]      Press CTRL+C to exit

    After about a minute, Azure IoT Akri issues the first discovery request based on the configuration:

    2023-06-07 12:49:17.344 +00:00 dbug: Grpc.AspNetCore.Server.ServerCallHandler[10]
      => SpanId:603279c62c9ccbb0, TraceId:15ad328e1e803c55bc6731266aae8725, ParentId:0000000000000000 => ConnectionId:0HMR7AMCHHG2G => RequestPath:/v0.DiscoveryHandler/Discover RequestId:0HMR7AMCHHG2G:00000001
      Reading message.
2023-06-07 12:49:18.046 +00:00 info: OpcUa.AssetDiscovery.Akri.Services.DiscoveryHandlerService[0]
      => SpanId:603279c62c9ccbb0, TraceId:15ad328e1e803c55bc6731266aae8725, ParentId:0000000000000000 => ConnectionId:0HMR7AMCHHG2G => RequestPath:/v0.DiscoveryHandler/Discover RequestId:0HMR7AMCHHG2G:00000001
      Got discover request opcuaDiscoveryMethod:
        - asset:
            endpointUrl: "opc.tcp://opcplc-000000:50000"
            useSecurity: false
            autoAcceptUntrustedCertificates: true
       from ipv6:[::ffff:10.1.7.47]:39708
2023-06-07 12:49:20.238 +00:00 info: OpcUa.AssetDiscovery.Akri.Services.DiscoveryHandlerService[0]
      => SpanId:603279c62c9ccbb0, TraceId:15ad328e1e803c55bc6731266aae8725, ParentId:0000000000000000 => ConnectionId:0HMR7AMCHHG2G => RequestPath:/v0.DiscoveryHandler/Discover RequestId:0HMR7AMCHHG2G:00000001
      Start asset discovery
2023-06-07 12:49:20.242 +00:00 info: OpcUa.AssetDiscovery.Akri.Services.DiscoveryHandlerService[0]
      => SpanId:603279c62c9ccbb0, TraceId:15ad328e1e803c55bc6731266aae8725, ParentId:0000000000000000 => ConnectionId:0HMR7AMCHHG2G => RequestPath:/v0.DiscoveryHandler/Discover RequestId:0HMR7AMCHHG2G:00000001
      Discovering OPC UA endpoint opc.tcp://opcplc-000000:50000 using Asset Discovery
...
2023-06-07 14:20:03.905 +00:00 info: OpcUa.Common.Dtdl.DtdlGenerator[6901]
      => SpanId:603279c62c9ccbb0, TraceId:15ad328e1e803c55bc6731266aae8725, ParentId:0000000000000000 => ConnectionId:0HMR7AMCHHG2G => RequestPath:/v0.DiscoveryHandler/Discover RequestId:0HMR7AMCHHG2G:00000001
      Created DTDL_2 model for boiler_1 with 35 telemetries in 0 ms
2023-06-07 14:20:04.208 +00:00 info: OpcUa.AssetDiscovery.Akri.CustomResources.CustomResourcesManager[0]
      => SpanId:603279c62c9ccbb0, TraceId:15ad328e1e803c55bc6731266aae8725, ParentId:0000000000000000 => ConnectionId:0HMR7AMCHHG2G => RequestPath:/v0.DiscoveryHandler/Discover RequestId:0HMR7AMCHHG2G:00000001
      Generated 1 asset CRs from discoveryUrl opc.tcp://opcplc-000000:50000
2023-06-07 14:20:04.208 +00:00 info: OpcUa.Common.Client.OpcUaClient[1005]
      => SpanId:603279c62c9ccbb0, TraceId:15ad328e1e803c55bc6731266aae8725, ParentId:0000000000000000 => ConnectionId:0HMR7AMCHHG2G => RequestPath:/v0.DiscoveryHandler/Discover RequestId:0HMR7AMCHHG2G:00000001
      Session ns=8;i=1828048901 is closing
...
2023-06-07 14:20:05.002 +00:00 info: OpcUa.AssetDiscovery.Akri.Services.DiscoveryHandlerService[0]
      => SpanId:603279c62c9ccbb0, TraceId:15ad328e1e803c55bc6731266aae8725, ParentId:0000000000000000 => ConnectionId:0HMR7AMCHHG2G => RequestPath:/v0.DiscoveryHandler/Discover RequestId:0HMR7AMCHHG2G:00000001
      Sending response to caller ...
2023-06-07 14:20:05.003 +00:00 dbug: Grpc.AspNetCore.Server.ServerCallHandler[15]
      => SpanId:603279c62c9ccbb0, TraceId:15ad328e1e803c55bc6731266aae8725, ParentId:0000000000000000 => ConnectionId:0HMR7AMCHHG2G => RequestPath:/v0.DiscoveryHandler/Discover RequestId:0HMR7AMCHHG2G:00000001
      Sending message.
2023-06-07 14:20:05.004 +00:00 info: OpcUa.AssetDiscovery.Akri.Services.DiscoveryHandlerService[0]
      => SpanId:603279c62c9ccbb0, TraceId:15ad328e1e803c55bc6731266aae8725, ParentId:0000000000000000 => ConnectionId:0HMR7AMCHHG2G => RequestPath:/v0.DiscoveryHandler/Discover RequestId:0HMR7AMCHHG2G:00000001
      Sent successfully

    After the discovery is completed, the result is sent back to Azure IoT Akri to create an Akri instance custom resource with asset information and observable variables. The discovery handler repeats the discovery every 10 minutes to detect changes on the server.

  4. To view the discovered Azure IoT Akri instances, run the following command:

    kubectl get akrii -n azure-iot-operations

    It might take a few minutes for the instance to show up.

    The output from the previous command looks like the following example. You might need to wait for a few seconds for the Akri instance to be created:

    NAMESPACE              NAME                      CONFIG             SHARED   NODES            AGE
azure-iot-operations   akri-opcua-asset-dbdef0   akri-opcua-asset   true     ["my-aio-vm"]   35m

    You can inspect the instance custom resource by using an editor such as OpenLens, under CustomResources/akri.sh/Instance.

    You can also view the custom resource definition YAML of the instance that was created:

    kubectl get akrii -n azure-iot-operations -o yaml

    The OPC UA Connector supervisor watches for new Azure IoT Akri instance custom resources of type opc-ua-asset, and generates the initial asset types and asset custom resources for them. You can modify asset custom resources to add settings such as extending publishing for more data points, or to add OPC UA Broker observability settings.

  5. To confirm that the Akri instance properly connected to the OPC UA Broker, copy and paste the name of the Akri instance that was the output of the previous step, into the following command:

    kubectl get akrii <AKRI_INSTANCE_NAME> -n azure-iot-operations -o json

    The command output looks like the following example. This example output shows the Akri instance brokerProperties values and confirms that the OPC UA Broker is connected.

    "spec": {

        "brokerProperties": {
            "ApplicationUri": "Boiler #2",
            "AssetEndpointProfile": "{\"spec\":{\"uuid\":\"opc-ua-broker-opcplc-000000-azure-iot-operation\"……

Related content