Urgent: Unable to Scale Compute (SKU) on Azure Database for PostgreSQL Flexible Server via Portal

Ícaro Sousa 20 Reputation points
2026-01-20T11:49:48.7133333+00:00

Hello,

I am requesting urgent assistance regarding a vertical scaling operation for my Azure Database for PostgreSQL Flexible Server instance.

I need to upgrade the compute tier. However, I am unable to complete this process through the Azure Portal interface. The scaling operation doesn't show any options.

Given the urgency of this performance upgrade, please investigate the issue or provide a backend method to apply this change immediately.

Current SKU: Standard_B1MS

Region: East US

Best regards,

Icaro Sousa

Azure Database for PostgreSQL
{count} votes

2 answers

Sort by: Most helpful
  1. Pilladi Padma Sai Manisha 3,875 Reputation points Microsoft External Staff Moderator
    2026-01-20T12:42:18.6633333+00:00

    Hi Ícaro Sousa
    Thankyou for reaching Microsoft Q&A!
    We understand the urgency of upgrading compute for your Azure Database for PostgreSQL Flexible Server and appreciate the details you have shared.

    Based on the information provided, this behavior is expected for your current configuration.

    Your server is running on the Standard_B1MS SKU, which belongs to the Burstable (B-series) compute tier. PostgreSQL Flexible Server does not support in-place vertical scaling from Burstable SKUs to other compute tiers. Because of this platform-level limitation, the Azure Portal does not present any scaling options, and the operation cannot be initiated via the portal, CLI, ARM, or any backend process.

    At this time, there is no backend or internal method available to force or manually apply a compute tier change for B-series servers. This restriction is enforced by the service architecture.

    Supported resolution:

    To upgrade performance, the supported approach is to move to a new server created with the required compute tier (for example, General Purpose or Memory Optimized):

    1. Create a new Azure Database for PostgreSQL Flexible Server in East US using the desired SKU.
    2. Migrate your data using one of the supported methods:
      • Azure Database Migration Service
      • pg_dump and pg_restore
      • Logical replication for minimal downtime scenarios
    3. Update your application connection strings to point to the new server.
    4. Validate functionality and performance before decommissioning the existing B1MS server.

    If future vertical scaling is expected, we recommend using General Purpose SKUs, which support seamless compute scaling without requiring server recreation.

    Please let us know if you would like assistance selecting an appropriate target SKU or guidance on the fastest migration path for your workload. We are happy to help you move forward as quickly as possible.

    0 comments No comments

  2. Q&A Assist
    2026-01-20T11:49:59.53+00:00
    AI generated content. This question contains an answer created with AI from Q&A Assist. Learn more

    To scale the compute tier of your Azure Database for PostgreSQL Flexible Server, you can use the Azure CLI as a workaround if the Azure Portal does not provide the necessary options. You can initiate the scaling operation using the following command:

    az postgres flexible-server update \
      --resource-group <resource_group> \
      --name <server> \
      --tier <tier> \
      --sku-name <sku_name>
    

    Make sure to replace <resource_group>, <server>, <tier>, and <sku_name> with your actual resource group name, server name, desired tier, and SKU name respectively. The allowed values for the --sku-name parameter depend on the tier you choose and the region.

    If you encounter an error regarding the SKU name, it may indicate that the SKU does not match the specified tier. Ensure you are selecting a valid SKU for the Standard tier you are trying to upgrade to.

    If you need to check the available SKU options for your region, refer to the documentation for the compute tiers and their specifications.


    References:


Your answer

Answers can be marked as 'Accepted' by the question author and 'Recommended' by moderators, which helps users know the answer solved the author's problem.