Edit

Share via

Pre-release testing

  • Article

To provide an opportunity for customers and partners to provide pre-release feedback, pre-release versions of Azure Arc-enabled data services are made available on a predictable schedule. This article describes how to install pre-release versions of Azure Arc-enabled data services and provide feedback to Microsoft.

Pre-release testing schedule

Each month, Azure Arc-enabled data services is released on the second Tuesday of the month, commonly known as "Patch Tuesday". The pre-release versions are made available on a predictable schedule in alignment with that release date.

  • 14 days before the release date, the test pre-release version is made available.
  • 7 days before the release date, the preview pre-release version is made available.

Normally, the main difference between the test and preview pre-release versions is quality and stability, but in some exceptional cases there may be new features introduced in between the test and preview releases.

Normally, pre-release version binaries are available around 10:00 AM Pacific Time. Documentation follows later in the day.

Artifacts for a pre-release version

Pre-release versions simultaneously release with artifacts, which are designed to work together:

  • Container images hosted on the Microsoft Container Registry (MCR)

    • mcr.microsoft.com/arcdata/test is the repository that hosts the test pre-release builds

    • mcr.microsoft.com/arcdata/preview is the repository that hosts the preview pre-release builds

      Note

      mcr.microsoft.com/arcdata/ will continue to be the repository that hosts the final release builds.

    • Azure CLI extension hosted on Azure Blob Storage

    • Azure Data Studio extension hosted on Azure Blob Storage

In addition to the above installable artifacts, the following are updated in Azure as needed:

  • New version of ARM API (occasionally)
  • New Azure portal accessible via a special URL query string parameter (see below for details)
  • New Arc-enabled Kubernetes extension version for Arc-enabled data services (applies to direct connectivity mode only)
  • Documentation updates on this page describing the location and details of the above artifacts and the new features available and any pre-release "read me" documentation

Installing pre-release versions

Install prerequisite tools

To install a pre-release version, follow these pre-requisite instructions:

If you use the Azure CLI extension:

  1. Uninstall the Azure CLI extension (az extension remove -n arcdata).
  2. Download the latest pre-release Azure CLI extension .whl file from the link in the Current preview release information.
  3. Install the latest pre-release Azure CLI extension (az extension add -s <location of downloaded .whl file>).

If you use the Azure Data Studio extension to install:

  1. Uninstall the Azure Data Studio extension. Select the Extensions panel and select on the Azure Arc extension, select Uninstall.
  2. Download the latest pre-release Azure Data Studio extension .vsix files from the links in the Current preview release information.
  3. Install the extensions. Choose File > Install Extension from VSIX package. Locate the download location of the .vsix files. Install the azcli extension first and then arc.

Install using Azure CLI

To install with the Azure CLI, follow the steps for your connectivity mode:

Indirect connectivity mode

  1. Set environment variables. Set variables for:

    • Docker registry
    • Docker repository
    • Docker image tag
    • Docker image policy

    Use the example script below to set environment variables for your respective platform.

    ## variables for the docker registry, repository, and image
export DOCKER_REGISTRY=<Docker registry>
export DOCKER_REPOSITORY=<Docker repository>
export DOCKER_IMAGE_TAG=<Docker image tag>
export DOCKER_IMAGE_POLICY=<Docker image policy>

  2. Follow the instructions to create a custom configuration profile.

  3. Use the command az arcdata dc create as explained in create a custom configuration profile.

Direct connectivity mode

If you install using the Azure CLI:

  1. Set environment variables. Set variables for:

    • Docker registry
    • Docker repository
    • Docker image tag
    • Docker image policy
    • Arc data services extension version tag (ARC_DATASERVICES_EXTENSION_VERSION_TAG): Use the version of the Arc enabled Kubernetes helm chart extension version from the release details under Current preview release information.
    • Arc data services release train: ARC_DATASERVICES_EXTENSION_RELEASE_TRAIN: { test | preview }.

    Use the example script below to set environment variables for your respective platform.

    ## variables for the docker registry, repository, and image
export DOCKER_REGISTRY=<Docker registry>
export DOCKER_REPOSITORY=<Docker repository>
export DOCKER_IMAGE_TAG=<Docker image tag>
export DOCKER_IMAGE_POLICY=<Docker image policy>
export ARC_DATASERVICES_EXTENSION_VERSION_TAG=<Version tag>
export ARC_DATASERVICES_EXTENSION_RELEASE_TRAIN='preview'

  2. Run az arcdata dc create as normal for the direct mode to:

    • Create the extension, if it doesn't already exist
    • Create the custom location, if it doesn't already exist
    • Create data controller

    For details see, create a custom configuration profile.

Install using Azure Data Studio

Note

Deploying pre-release builds using direct connectivity mode from Azure Data Studio is not supported.

You can install with Azure Data Studio (ADS) in indirect connectivity mode. To use Azure Data Studio to install:

  1. Complete the data controller deployment wizard as normal except click on Script to notebook at the end instead of Deploy.

  2. Update the following script. Replace { test | preview } with the appropriate label.

  3. In the generated notebook, edit the Set variables cell to add the following lines:

    # choose between arcdata/test or arcdata/preview as appropriate
os.environ["AZDATA_DOCKER_REPOSITORY"] = "{ test | preview }"
os.environ["AZDATA_DOCKER_TAG"] = "{ Current preview tag }

  4. Run the notebook, click Run All.

Install using Azure portal

  1. Follow the instructions to Arc-enabled the Kubernetes cluster as normal.

  2. Open the Azure portal for the appropriate preview version:

  3. Follow the instructions to Create the Azure Arc data controller from Azure portal - Direct connectivity mode except that when choosing a deployment profile, select Custom template in the Kubernetes configuration template drop-down.

  4. Set the repository to either arcdata/test or arcdata/preview as appropriate. Enter the desired tag in the Image tag field.

  5. Fill out the rest of the custom cluster configuration template fields as normal.

Complete the rest of the wizard as normal.

When you deploy with this method, the most recent pre-release version will always be used.

Current preview release information

At this time, a test or preview build is not available for the next release.

Provide feedback

At this time, pre-release testing is supported for certain customers and partners that have established agreements with Microsoft. Participants have points of contact on the product engineering team. Email your points of contact with any issues that are found during pre-release testing.

Related content

Release notes - Azure Arc-enabled data services