What Microsoft has said about the tiers
Microsoft's "What's New" page for Azure Cache for Redis gives two separate April 2026 entries:
- Tooling to aid migration from Azure Cache for Redis Enterprise is now generally available. When you start a migration with it, the Enterprise instance is transformed into an Azure Managed Redis instance.
- Tooling for Basic, Standard, and Premium was announced in Public Preview. That tooling updates the cache endpoint to point to a precreated Azure Managed Redis instance.
The Azure Updates page I could retrieve didn't show the announcement text. It returned no update record. So I can't say whether the November 17 item covers Enterprise only or extends to Basic, Standard and Premium. The Learn pages I could see still label the Basic/Standard/Premium workflow as preview. Check the exact tier wording in the announcement and the current documentation before treating the whole fleet as covered by a GA SLA.
Retirement dates
The timeline explains why this tooling matters:
- Microsoft tells customers to migrate all instances to Azure Managed Redis by September 30, 2028.
- That date applies to Basic, Standard and Premium. In Azure Government and Azure operated by 21Vianet, the same date applies, with no creation block before it.
- In July 2026, Microsoft dropped the planned October 2026 creation block for existing customers. A tenant counts as existing if any of its subscriptions had an Azure Cache for Redis instance before April 1, 2026.
- Enterprise and Enterprise Flash timelines are unchanged. An earlier schedule on the same page said remaining Enterprise-tier caches would be migrated to Azure Managed Redis on March 31, 2027.
How the tooling works
For Enterprise, the Enterprise hostname points to an Azure Managed Redis instance after migration. Client applications reconnect automatically with the same hostname and access key. Once you validate the migration, update clients to the new Azure Managed Redis hostname. Microsoft says the old hostname will be automatically deleted in future, so don't treat the redirect as permanent.
For Basic, Standard and Premium (the preview workflow), the documented steps are:
- Provision the target first. Update your deployment scripts, such as ARM templates, Bicep files or Terraform configurations, to provision Azure Managed Redis. Pick a size that is the same as or larger than the existing cache, using Microsoft's SKU mapping table.
- Set up Entra ID if you use it. Configure the required managed identities and permissions on the target before you migrate.
- Copy data separately if you need it. Microsoft lists data migration as an optional step with its own strategies. One independent write-up, from a blogger rather than Microsoft, makes the same point. It says the tooling handles hostname cutover only and does not migrate data.
- Validate, then migrate. In the portal, open the source cache's Overview and select Migrate. Choose the target and run Validate. Validation can return warnings or errors. Persistence enabled on the source but not the target is a warning. A cache injected into a virtual network is an error, because that setup is unsupported. You can bypass warnings after reviewing them.
- Expect a blip. The cache status shows Migrating, and no other management operations are possible until it finishes. Clients see a connection blip similar to a maintenance event, then reconnect to Azure Managed Redis.
- Validate, delete, then repoint. Test the application, then delete the old cache. The old hostname keeps pointing at the new instance even after deletion. Move applications to the new hostname, formatted
<cachename>.<region>.redis.azure.net. Microsoft says an API to retire the old hostname is on its roadmap. - Script it if you prefer. Microsoft also documents a PowerShell route for pre-validation, starting the migration, checking status and cancelling.
Practical advice
- Treat it as a cutover tool. The tooling redirects the endpoint. It doesn't promise zero downtime, and it doesn't copy your cache contents. If your application can't tolerate a cold cache, plan data migration separately.
- Run validation early. Virtual-network injection is a hard blocker for the Basic/Standard/Premium workflow. Find that out in a test window, not on cutover day.
- Pre-stage identity and networking. Entra ID permissions must be on the target before you start. If you use private endpoints, the target also needs matching network setup.
- Don't coast on the hostname redirect. The old name is a stopgap. Schedule the application change and track it.
- Know which feature caveats apply. Azure Managed Redis became GA as a product in May 2025. Some features were still in preview then, including scaling, data persistence, non-clustered caches, in-memory tiers above 120 GB, and all Flash Optimized tiers. Check the current status of any feature you depend on. The migration announcement doesn't change those statuses.
Bottom line
A GA label for migration tooling is good news for anyone with a 2027–2028 retirement deadline. Whether it applies to your tier depends on wording I couldn't see. Microsoft's documentation confirms GA for the Enterprise tooling and keeps the Basic/Standard/Premium tooling in preview. Confirm which applies to you, run validation in a non-production environment, and keep data migration and hostname cleanup on your plan.
References
- (Launched) Generally Available: Azure Cache for Redis to Azure Managed Redis migration tooling Azure Updates · 2026-11-17T20:00:32Z
- Migrate with tooling - Migrate from Basic, Standard, and Premium tiers to Azure Managed Redis - Azure Managed Redis | Microsoft Learn learn.microsoft.com
- What's New in Azure Cache for Redis - Azure Cache for Redis learn.microsoft.com