Edit

Share via

Known issues: Azure IoT Operations

This article lists the current known issues you might encounter when using Azure IoT Operations. The guidance helps you identify these issues and provides workarounds where available.

For general troubleshooting guidance, see Troubleshoot Azure IoT Operations.

Deploy, update, and uninstall issues

This section lists current known issues that might occur when you deploy, update, or uninstall Azure IoT Operations.

Error creating custom resources

Issue ID: 9091

Log signature: "code": "ExtensionOperationFailed", "message": "The extension operation failed with the following error: Error occurred while creating custom resources needed by system extensions"

The message Error occurred while creating custom resources needed by system extensions indicates that your deployment failed due to a known sporadic issue.

To work around this issue, use the az iot ops delete command with the --include-deps flag to delete Azure IoT Operations from your cluster. When Azure IoT Operations and its dependencies are deleted from your cluster, retry the deployment.

Codespaces restart error

Issue ID: 9941

Log signature: "This codespace is currently running in recovery mode due to a configuration error."

If you deploy Azure IoT Operations in GitHub Codespaces, shutting down and restarting the Codespace causes a This codespace is currently running in recovery mode due to a configuration error issue.

There's no workaround for this issue. If you need a cluster that supports shutting down and restarting, select one of the options in Prepare your Azure Arc-enabled Kubernetes cluster.

MQTT broker issues

This section lists current known issues for the MQTT broker.

MQTT broker resources aren't visible in the Azure portal

Issue ID: 4257

Log signature: N/A

MQTT broker resources created in your cluster using Kubernetes aren't visible in the Azure portal. This result is expected because managing Azure IoT Operations components using Kubernetes is in preview, and synchronizing resources from the edge to the cloud isn't currently supported.

There's currently no workaround for this issue.

Azure IoT Layered Network Management (preview) issues

This section lists current known issues for Azure IoT Layered Network Management.

Layered Network Management service doesn't get an IP address

Issue ID: 7864

Log signature: N/A

The Layered Network Management service doesn't get an IP address when it runs on K3S on an Ubuntu host.

To work around this issue, you reinstall K3S without the traefik ingress controller:

curl -sfL https://get.k3s.io | sh -s - --disable=traefik --write-kubeconfig-mode 644

To learn more, see Networking | K3s.

CoreDNS service doesn't resolve DNS queries correctly

Issue ID: 7955

Log signature: N/A

DNS queries don't resolve to the expected IP address while using the CoreDNS service running on the child network level.

To work around this issue, upgrade to Ubuntu 22.04 and reinstall K3S.

Connector for OPC UA issues

This section lists current known issues for the connector for OPC UA.

Connector pod doesn't restart after configuration change

Issue ID: 7518

Log signature: N/A

When you add a new asset with a new asset endpoint profile to the OPC UA broker and trigger a reconfiguration, the deployment of the opc.tcp pods changes to accommodate the new secret mounts for username and password. If the new mount fails for some reason, the pod doesn't restart and therefore the old flow for the correctly configured assets stops as well.

An OPC UA server modeled as a device can only have one inbound endpoint of type "Microsoft.OpcUa"

Issue ID: 2411

2025-07-24T13:29:30.280Z aio-opc-supervisor-85b8c78df5-26tn5 - Maintaining the new asset test-opcua-asset | - | 1 is skipped because the endpoint profile test-opcua.opcplc-e2e-anon-000000 is not present

When you create an OPC UA device, you can only have one inbound endpoint of type Microsoft.OpcUa. Currently, any other endpoints aren't used.

Workaround: Create multiple devices with a single endpoint each if you want to use namespace assets.

An OPC UA namespaced asset can only have a single dataset. Currently, any other datasets aren't used.

Workaround: Create multiple namespace assets each with a single dataset.

Data spike every 2.5 hours with some OPC UA simulators

Issue ID: 6513

Log signature: Increased message volume every 2.5 hours

Data values spike every 2.5 hours when using particular OPC UA simulators causing CPU and memory spikes. This issue isn't seen with OPC PLC simulator used in the quickstarts. No data is lost, but you can see an increase in the volume of data published from the server to the MQTT broker.

Connector for media and connector for ONVIF issues

This section lists current known issues for the connector for media and the connector for ONVIF.

AssetType CRD removal process doesn't complete

Issue ID: 6065

Log signature: "Error HelmUninstallUnknown: Helm encountered an error while attempting to uninstall the release aio-118117837-connectors in the namespace azure-iot-operations. (caused by: Unknown: 1 error occurred: * timed out waiting for the condition"

Sometimes, when you attempt to uninstall Azure IoT Operations from the cluster, the system reaches a state where CRD removal job is stuck in pending state, which blocks the cleanup of Azure IoT Operations.

To work around this issue, complete the following steps to manually delete the CRD and finish the uninstall:

  1. Delete the AssetType CRD manually: kubectl delete crd assettypes.opcuabroker.iotoperations.azure.com --ignore-not-found=true

  2. Delete the job definition: kubectl delete job aio-opc-delete-crds-job-<version> -n azure-iot-operations

  3. Find the Helm release for the connectors, it's the one with -connectors suffix: helm ls -a -n azure-iot-operations

  4. Uninstall Helm release without running the hook: helm uninstall aio-<id>-connectors -n azure-iot-operations --no-hooks

Media and ONVIF devices with an underscore character in the endpoint name are ignored

Issue ID: 5712

Log signature: N/A

If you create a media or ONVIF device with an endpoint name that contains an underscore ("_") character, the connector for media ignores the device.

To work around this issue, use a hyphen ("-") instead of an underscore in the endpoint name.

Media connector doesn't use the path in destination configuration

Issue ID: 6797

Log signature: N/A

Media assets with a task type of "snapshot-to-fs" or "clip-to-fs" don't honor the path in the destination configuration. Instead, they use the "Additional configuration" path field.

Media connector ignores MQTT topic setting in asset

Issue ID: 6780

Log signature: N/A

The media connector ignores the MQTT destination topic setting in the asset. Instead, it uses the default topic: /azure-iot-operations/data/<asset-name>/snapshot-to-mqtt.

Media connector inbound endpoint addresses aren't fully validated

Issue ID: 2679

Log signature: N/A

In the public preview release, the media connector accepts device inbound endpoint addresses with the following schemes: async, cache, concat, concatf, crypto, data, fd, ffrtmpcrypt, ffrtmphttp, file, ftp, gopher, gophers, hls, http, httpproxy, https, mmsh, mmst, pipe, rtmp, rtmpe, rtmps, rtmpt, rtmpte, rtmpts, rtp, srtp, subfile, tcp, tls, udp, udplite, unix, ipfx, ipns.

This enables input data from multiple source types. However, because the output configuration is based on the streamConfiguration, the possibilities for using data from these sources are limited.

Secret sync conflict

Issue ID: 0606

Log signature: N/A

When using secret sync, ensure that secret names are globally unique. If a local secret with the same name exists, connectors might fail to retrieve the intended secret.

Data flows issues

This section lists current known issues for data flows.

Data flow resources aren't visible in the operations experience web UI

Issue ID: 8724

Log signature: N/A

Data flow custom resources created in your cluster using Kubernetes aren't visible in the operations experience web UI. This result is expected because managing Azure IoT Operations components using Kubernetes is in preview, and synchronizing resources from the edge to the cloud isn't currently supported.

There's currently no workaround for this issue.

A data flow profile can't exceed 70 data flows

Issue ID: 0313

Log signature:

exec /bin/main: argument list too long

If you create more than 70 data flows for a single data flow profile, deployments fail with the error exec /bin/main: argument list too long.

To work around this issue, create multiple data flow profiles and distribute the data flows across them. Don't exceed 70 data flows per profile.

The request persistence flag is not set for MQTT sessions created for data flow graphs (WASM)

Issue ID: 6415

Log signature: N/A

When you create a data flow graph using the WASM, the MQTT session doesn't have the request persistence flag set.

To work around this issue, set MQTT broker Retained messages mode to All. For more information, see Configure MQTT broker persistence.

Complex data might be flattened when enriching data in a data flow

Issue ID: 7385

Log signature: N/A

When enriching data using complex object DSS reference data, the output might be moved to the root level instead of preserving the original structure.

For example, if you have a complex object with properties like:

{
  "complex_property_1": {
      "field1": 12,
      "field2": 13 
  },
  "complex_property_2": {
     "field2": 24
  }
}

The output might look like:

{
  "property_1": 2,
  "property_2": 3,
  "field1": 12,
  "field2": 24,
}

The complex properties are flattened to the root, and the original structure is lost. If fields with the same name exist in the complex objects or the root, the values might overwrite the root values. In the example, field2 from complex_property_2 overwrites the field2 from complex_property_1.