Batch Support for Windows Server 2016 Marketplace images ending

Azure Batch periodically retires support for operating systems that reach end-of-life status. Support for [Windows Server 2016](https://learn.microsoft.com/lifecycle/products/windows-server-2016) concluded on 11 January 2022, with extended support scheduled to end on **12 January 2027**. Marketplace images for Windows Server 2016, published under the _MicrosoftWindowsServer_ offer with the name _WindowsServer_, will no longer be available. The relevant Marketplace image SKUs include:

- 2016-datacenter-gensecond
- 2016-datacenter
- 2016-datacenter-gs
- 2016-datacenter-smalldisk
- 2016-datacenter-smalldisk-g2
- 2016-datacenter-with-containers
- 2016-datacenter-with-containers-gs

If your workload still uses Batch pools based on these Windows Server 2016 Marketplace images, migrate your workload to a supported Windows Server image no later than **12 January 2027**.

If your workload requires Windows Server 2016 after **12 January 2027**, you may be able to migrate to a Windows Server 2016 custom image with appropriate licensing and Azure ESU. Please visit this [article](https://learn.microsoft.com/azure/batch/batch-sig-images) for more information about this option. However, Azure Batch support for Windows Server 2016 will end on **12 January 2027**. We recommend upgrading to the latest version of Windows Server, if possible.

As a reminder, we recommend reviewing the [Batch best practices guide](https://learn.microsoft.com/azure/batch/best-practices) which outlines support expectations, recommendations on Batch pool image freshness, and how to query for Batch image support at end-of-life dates.

### Required action

To avoid service disruptions, migrate your Batch pools with affected Marketplace images to a supported image before **12 January 2027**. Please refer to the article on how to [update pool properties](https://learn.microsoft.com/azure/batch/batch-pool-update-properties) for migration guidance. We recommend testing the migration before updating production configurations.