Introduction
Azure AI Search is announcing two new preview features that make it easier to scale your search service to avoid production issues from storage limitations as your needs grow.
Available in preview today:
- Change your pricing tier: Change the tier of your existing Azure AI Search service via the portal or management plane REST API.
- Self-service upgrade: Upgrade your search service to enable features previously only available in new services, such as the new storage limits released in April 2024.
Change your pricing tier to scale up
Now you can change your service tier from the Azure portal or management plane API. It's a simple scaling operation like adding partitions or replicas that ensures uninterrupted growth and operational continuity.
Before, if you reached the maximum total storage limit or vector index size supported by your pricing you had to either delete data, create a new service at a higher tier and reindex, or open a support case. These options were time consuming and difficult to complete on a production service.
This has been the most popular feature request from Azure AI Search customers, and we’re very happy to make it available to you.
This screenshot shows the Scale pane in the Azure portal for anAzure AI Search service. It highlights a new feature where you can now change the Pricing tier of an existing search service.
In the portal, just click Settings->Scale, select a new pricing tier, click Save, and you’re done! The process is as easy as adding partitions and replicas. Your service is available for use while the operation is running.
More information:
- You can change between the Basic, S1, S2, and S3 pricing tiers.
- Changing the tier of an S3 HD, L1, or L2 service is not supported.
- A service on the Free tier can’t be changed to any of the paid service tiers.
- Please note: At launch you can only increase your pricing tier. You will be able to lower your pricing tier in a future update.
Increased storage capacity for existing services
Some Azure AI Search features can only be released in new search services. In April 2024, we announced a 5x to 6x increase in the total storage capacity and vector index size of new search services at no additional cost. With the preview release of service upgrade, you can easily bring these higher storage limits to existing search services without having to recreate your service. You may also be able to reduce storage costs by fitting your data into fewer partitions.
AT&T, an Azure AI Search customer, can scale their RAG platform without compromising cost by upgrading their services to higher storage limits.
See the Azure AI Search documentation to learn more.
The screenshot shows the Overview pane of theAzure portal for an Azure AI Search service. It highlights the new Upgrade option in the command bar. This option is only enabled if an upgrade is available for your service.In the portal, the Upgrade button in the command bar of the Overview pane will be enabled if an upgrade is currently available for your search service. Click Upgrade, confirm the operation, and you’re all set! The upgrade operation will begin asynchronously. Just like other scale operations, your service will continue to be available. Once the upgrade is complete, you will see higher values for both total storage and vector index quota on the Usage tab.
More information:
- Higher storage limits is the only upgrade currently available. Check the portal to see if an upgrade is currently available for your service.
- Additional upgrades may be available in the future. Look for updates in the Azure AI Search documentation.
- Service upgrades are available in select regions at launch. Additional regions will be supported in a future release.
Getting started with Azure AI Search
- Learn about Azure AI Search and the latest features
- Start creating a search service in the Azure Portal, Azure CLI, the Management REST API, ARM template, or a Bicep file.
- Learn about higher storage limits: Announcing cost-effective RAG at scale with Azure AI Search
- Learn about Retrieval Augmented Generation in Azure AI Search.
- Explore how to create end-to-end RAG applications with Azure AI Foundry.