Azure scale deployment freeze

Question

Azure scale deployment freeze

Andrii K 0

Hello,
Azure scale deployment freeze in the loading state. No possibilities to cancel or delete it.

Support request is not available for a Basic plan.
This is the issue on the side of Azure. How it can be resolved without the support request? User's image

Andrii K 0 Reputation points

2025-05-30T00:32:04.4733333+00:00

Hello Prasad Chaganti,

Thank you for your perfect guide. It's much better than a support responses.

I wasn't able to delete deployment because of some "previous deployment" that is not showing in the deployments list.

I tried to cancel it and was able to delete it after.
However, the server is still in "Updating" state.

I tried to apply storage update from cli, but I had a response that server is busy with another operation.
Andrii K 0 Reputation points

2025-05-30T15:02:16.04+00:00

@Prasad Chaganti , do you know how I can clean that "zombie" deployment that is not showing?
Prasad Chaganti 770 Reputation points Microsoft External Staff Moderator

2025-05-31T00:27:09.7366667+00:00
Hi Andrii K,

You're encountering a persistent issue where your Azure server is stuck in the "Updating" state, and any attempt to apply changes (like a storage update) fails with a message that the server is busy with another operation. This is a known problem in Azure, particularly with Flexible Server and SQL Managed Instance resources

Why the Server Is Stuck

This typically occurs when:

A previous deployment or update operation failed or was interrupted.

The Azure control plane still believes an operation is in progress.

The server is locked in a transitional state (e.g., "Updating", "Stopping", "Deleting").

This can block all further operations, including CLI commands, portal actions, and even deletion attempts.

What You Can Try Without a Support Plan

1.Check the Activity Log

Go to the Activity Log in the Azure Portal for the resource group or server:

Look for any failed operations or stuck deployments.

Identify the operation ID or correlation ID if available.

2.Use Azure CLI to Check Provisioning State

Run:

az resource show --ids <resource-id> --query "properties.provisioningState"

This will confirm if the server is truly stuck in Updating.

3.Force Retry or Cancel the Operation

If the resource is part of a deployment stack:

Use:

az stack group update --name <stack-name> --resource-group <rg-name> --detach-resources

Then try deleting or updating the server again.

If not part of a stack, try:

az sql server update --name <server-name> --resource-group <rg-name> --set tags.forceUpdate=true

This sometimes forces a metadata refresh.

4.Redeploy a Minimal Template

If the server is part of an ARM or Bicep deployment:

Redeploy a minimal or empty template to the same resource.

This can sometimes reset the provisioning state.

To clean up a "zombie" Azure deployment that is stuck in a transitional state (like "Updating") and not visible in the portal’s deployment list

1.Use Azure CLI to List and Cancel Hidden Deployments

Sometimes deployments are not visible in the portal but still exist in the backend. Use the following CLI command to list all deployments in the resource group:

az deployment group list --resource-group <your-resource-group-name> --output table

If you find a stuck deployment, cancel it:

az deployment group cancel --name <deployment-name> --resource-group <your-resource-group-name>

If it's a subscription-level deployment:

az deployment sub cancel --name <deployment-name>

2.Force Detach Resources from Deployment Stack

If the resource is part of a deployment stack (especially if created via Bicep or ARM templates), try detaching it:

az stack group update --name <stack-name> --resource-group <your-resource-group-name> --detach-resources

Then delete the stack:

az stack group delete --name <stack-name> --resource-group <your-resource-group-name>

3.Redeploy a Minimal Template

Deploying a minimal or empty ARM/Bicep template to the same resource can sometimes reset the provisioning state:

az deployment group create \ --resource-group <your-resource-group-name> \ --template-file empty-template.json

Where empty-template.json contains:

{ "$schema": "https://schema.management.azure.com/schemas/2019-04-01/deploymentTemplate.json#", "contentVersion": "1.0.0.0", "resources": [] }

4.Check Activity Logs for Hidden Operations

Go to Azure Portal > Monitor > Activity Log and filter by:

Resource group

Status: Failed or In Progress

Operation: Microsoft.Resources/deployments/write

This can help you identify the stuck deployment and its correlation ID.

I hope this information helps. Please do let us know if you have any further queries.

If the comment is helpful, please click "Upvote it".
Andrii K 0 Reputation points

2025-06-01T14:34:53.2266667+00:00
Thank you, @Prasad Chaganti !

Activity log from Monitor shows my previous activity, however, deployments were deleted before (2 failed and 1 running).

az resource show --ids <resource-id> --query "properties.provisioningState"
Shows nothing.

3.Server update is not working because of "performing previous operation"

Azure cli is not showing any additional failed/active deployments.

5.Empty deployment was executed successfully, but the server state persist the same "Updating".

6.Sub level deployment are empty.

7.No stacks available.

Do I have any additional options?
Deleted

This comment has been deleted due to a violation of our Code of Conduct. The comment was manually reported or identified through automated detection before action was taken. Please refer to our Code of Conduct for more information.
Andrii K 0 Reputation points

2025-06-03T13:22:16.0133333+00:00

What should I do next?
Database is down for 5 days.
Prasad Chaganti 770 Reputation points Microsoft External Staff Moderator

2025-06-04T11:31:58.2166667+00:00
Hi Andrii K,

your Azure database has been down for 5 days, and you're stuck in a persistent "Updating" state with no visible deployments or stacks

Confirm the Server Is Truly Stuck

az resource show --ids <resource-id> --query "properties.provisioningState"

If this returns nothing, it confirms the metadata is broken or inaccessible.

Clean Up Hidden or Zombie Deployments

List All Deployments in the Resource Group

az deployment group list --resource-group <your-rg-name> --output table

If you find any stuck deployments:

az deployment group cancel --name <deployment-name> --resource-group <your-rg-name>

Check for Subscription-Level Deployments

az deployment sub list --output table az deployment sub cancel --name <deployment-name>

Detach from Deployment Stack (If Applicable)

Even if no stacks are visible in the portal, try:

az stack group update --name <stack-name> --resource-group <your-rg-name> --detach-resources az stack group delete --name <stack-name> --resource-group <your-rg-name>

If this fails, it confirms the stack is locked or corrupted.

Redeploy a Minimal Template to Reset State

Create a file called empty-template.json:

{ "$schema": "https://schema.management.azure.com/schemas/2019-04-01/deploymentTemplate.json#", "contentVersion": "1.0.0.0", "resources": [] }

Then run:

az deployment group create \ --resource-group <your-rg-name> \ --template-file empty-template.json

This can sometimes overwrite the stuck state.

Force Metadata Refresh

If it’s a SQL server or Flexible Server:

az sql server update --name <server-name> --resource-group <rg-name> --set tags.forceUpdate=true

This may trigger a backend refresh.

Check Activity Logs for Hidden Operations

Go to: Azure Portal → Monitor → Activity Log

Filter by:

Resource group

Status: Failed or In Progress

Operation: Microsoft.Resources/deployments/write

Look for correlation IDs or operation IDs that might help identify the stuck process.

I hope this information helps. Please do let us know if you have any further queries.

If the comment is helpful, please click "Upvote it".
Andrii K 0 Reputation points

2025-06-04T12:43:52.9166667+00:00
Hello @Prasad Chaganti ,

az resource show --ids <resource-id> --query "properties.provisioningState"

Returns no data.

All other steps were performed previously. No deployments. No stacks.

Force update response:
Message: Server server_name is busy with other operations. Please try later

Sub list is empty.

New empty deployment runs with no issues, but changes nothing with the "updating" state.

None of the proposed solutions helped to fix the issue.
Prasad Chaganti 770 Reputation points Microsoft External Staff Moderator

2025-06-04T16:24:25.75+00:00
Hi Andrii K,

Since you've already exhausted the standard CLI and portal-based recovery options and the server remains stuck in the "Updating" state with no visible deployments, stacks, or operations. Please follow the below steps

Force a Metadata Refresh via Tag Mutation

Even if --set tags.forceUpdate=true failed, try changing a different tag or adding a new one:

az resource update --ids <resource-id> --set tags.recoveryAttempt=$(date +%s)

This can sometimes trigger a backend refresh even when the resource is stuck.

Try a PATCH via REST API

Use the Azure REST API directly to send a PATCH request to the resource with a minimal payload. This bypasses some CLI-level validations.

Example using curl:

curl -X PATCH \ -H "Authorization: Bearer <access-token>" \ -H "Content-Type: application/json" \ https://management.azure.com/<resource-id>?api-version=2021-04-01 \ -d '{"tags": {"forcePatch": "true"}}'

You can get the access token using:

az account get-access-token --query accessToken -o tsv

Try a Move Operation

Sometimes moving the resource to another resource group can reset its state:

az resource move --destination-group <new-rg> --ids <resource-id>

If this fails, it confirms the control plane is locked, but if it succeeds, it may reset the provisioning state.

Redeploy a Conflicting Resource

If the resource is a server (e.g., PostgreSQL Flexible Server), try deploying a new server with the same name in the same region and resource group. Azure may prompt you to resolve the conflict, which can sometimes clear the stuck state.

Use Azure Resource Explorer

Go to https://resources.azure.com, navigate to the resource, and attempt a manual PATCH or DELETE from there. This tool gives you low-level access to the resource provider API.

Could you please follow the steps above and let us know if they worked or not? If you have any queries, please do let us know we will try another way to resolve the issue.
Andrii K 0 Reputation points

2025-06-05T21:36:54.45+00:00

Hello @Prasad Chaganti ,

I was able to resolve the issue only with Microsoft Support help.

1 answer

Your answer

Andrii K 0 Reputation points

2025-05-30T00:32:04.4733333+00:00

Hello Prasad Chaganti,

Thank you for your perfect guide. It's much better than a support responses.

I wasn't able to delete deployment because of some "previous deployment" that is not showing in the deployments list.

I tried to cancel it and was able to delete it after.
However, the server is still in "Updating" state.

I tried to apply storage update from cli, but I had a response that server is busy with another operation.
Andrii K 0 Reputation points

2025-05-30T15:02:16.04+00:00

@Prasad Chaganti , do you know how I can clean that "zombie" deployment that is not showing?
Andrii K 0 Reputation points

2025-06-01T14:34:53.2266667+00:00

Thank you, @Prasad Chaganti !

Activity log from Monitor shows my previous activity, however, deployments were deleted before (2 failed and 1 running).

az resource show --ids <resource-id> --query "properties.provisioningState"
Shows nothing.

3.Server update is not working because of "performing previous operation"

Azure cli is not showing any additional failed/active deployments.

5.Empty deployment was executed successfully, but the server state persist the same "Updating".

6.Sub level deployment are empty.

7.No stacks available.

Do I have any additional options?
Deleted

This comment has been deleted due to a violation of our Code of Conduct. The comment was manually reported or identified through automated detection before action was taken. Please refer to our Code of Conduct for more information.
Andrii K 0 Reputation points

2025-06-03T13:22:16.0133333+00:00

What should I do next?
Database is down for 5 days.
Andrii K 0 Reputation points

2025-06-04T12:43:52.9166667+00:00

Hello @Prasad Chaganti ,

az resource show --ids <resource-id> --query "properties.provisioningState"

Returns no data.

All other steps were performed previously. No deployments. No stacks.

Force update response:
Message: Server server_name is busy with other operations. Please try later

Sub list is empty.

New empty deployment runs with no issues, but changes nothing with the "updating" state.

None of the proposed solutions helped to fix the issue.
Andrii K 0 Reputation points

2025-06-05T21:36:54.45+00:00

Hello @Prasad Chaganti ,

I was able to resolve the issue only with Microsoft Support help.

Answer 1

Hi Andrii K,

This issue typically occurs when:

A deployment stack or scale set operation fails mid-way.
The Azure control plane is left in an inconsistent state.
The deployment metadata is locked, preventing deletion or rollback.

1.Use Azure CLI to Force Delete the Deployment StackIf the deployment is stuck and cannot be deleted via the portal, try using the CLI:

az deployment sub delete --name <deployment-name>

Or for resource group deployments:

az deployment group delete --resource-group <rg-name> --name <deployment-name>

If the deployment is part of a deployment stack, try:

az stack sub delete --name <stack-name>

Note: If the stack is locked, this may still fail. In that case, proceed to the next step.

2.Detach Resources from the Stack

If the stack is blocking deletion due to attached resources:

Go to the Deployment Stack in the portal.
Choose Update behavior → Detach resources.
Then try deleting the stack again.

3.Then try deleting the stack again.

Sometimes re-deploying the same stack with a minimal or empty template can reset the state:

Use a Bicep or ARM template with no resources.
Deploy it to the same stack name.
This may allow the system to overwrite the stuck state and clean up.

4.Retry the Deployment via GitHub Action or DevOps Pipeline

In some cases, simply re-running the deployment pipeline (e.g., GitHub Actions or Azure DevOps) has resolved the issue by forcing the control plane to re-evaluate the deployment state.

I hope this information helps. Please do let us know if you have any further queries.

If the comment is helpful, please click "Upvote it".

Share via

Azure scale deployment freeze

1 answer

Your answer