Unable to provision new Azure Managed Redis instance

Question

Unable to provision new Azure Managed Redis instance

Suresh Panthangi 21

We are unable to provision new Azure Managed Redis instances. Existing Azure Managed Redis instances created approximately two weeks ago continue to operate normally. New deployments fail consistently from both the Azure Portal and Terraform (AzAPI). The deployment passes ARM validation but reaches terminal provisioning state Failed after approximately two minutes with OperationFailed. This occurs across multiple resource groups, resource names, deployment methods, and SKUs. Please investigate the backend provisioning logs using the provided Correlation ID, Tracking ID, and Service Request ID.

0 comments

1 answer

Your answer

Answer 1

TP 160.7K Volunteer Moderator

Hi Suresh,

Recently when I've seen that error it is due to lack of capacity in the region. You may have better luck in other regions, especially one that is not as popular. Unfortunately errors due to lack of capacity seem to be happening more frequently.

Another option you can try is to create a free Service and subscription limits (quotas) support request and ask for enough capacity in your preferred region.

To start this process you may use link below:

https://portal.azure.com/#create/Microsoft.Support

Issue type: Service and subscription limits (quotas)

Quota type: Other Requests

Click Next

On Additional details tab, explain you are getting capacity errors with Azure Managed Redis and detail capacity you need along with your preferred region(s) you would consider. Fill out remaining information, click Next, review and Create.

There are no guarantees they will approve your request.

Please click Accept Answer and upvote if the above was helpful.

Thanks.

-TP

Suresh Panthangi 21 Reputation points

2026-07-24T10:52:16.5633333+00:00
Hi TP,

Thank you for the suggestion. I have reproduced this issue:

Using Azure Portal and Terraform (AzAPI)

Across multiple resource groups

With different resource names

With different SKUs (including Consumption)

In multiple Azure regions

Existing Azure Managed Redis instances created two weeks ago continue to run successfully.

Every new deployment reaches provisioning for approximately two minutes and then fails with OperationFailed.

Does this still indicate a capacity issue.?
TP 160.7K Reputation points Volunteer Moderator

2026-07-27T08:52:35+00:00

I can't say 100% if it is capacity issue, but several other people have had same error recently and that was the underlying cause. If you tried like 30-40 different regions, and in each you tried the different SKUs, all with same result, then I would have different opinion. I say this because in my experience if I try different regions and different SKUs I can usually succeed at least in some cases.

I recommend creating free quota support case as I explained above and see if they will approve capacity for your needs or at least recommend different region with more available.
Ganesh Chelluri 90 Reputation points Microsoft External Staff Moderator

2026-08-02T07:37:51.11+00:00

Hi @Suresh Panthangi ,

Based on what you've described, this doesn't look like capacity. When it's capacity you normally get AllocationFailed with a message telling you to retry with a different size or region. You're getting plain OperationFailed, and you're getting it across multiple regions, multiple resource groups, and multiple SKUs including Consumption. A regional shortage wouldn't follow you around like that. Your deployment also passes validation and then dies at roughly the same two minute mark every time, which points at something failing in the provisioning step rather than the platform refusing to allocate. And your instances from two weeks back are still running fine, so something changed after they were created.

Couple of things that would help narrow it down. Post the full error body rather than just the top level code expand details/innerError on the failed operation. In the portal it's Resource group, Deployments, pick the failed one, Operation details, Raw Error. Also worth confirming the provider is still fine with az provider show -n Microsoft.Cache --query registrationState, and checking whether any subscription policy or deny assignment landed in the last couple of weeks.

On the IDs you mentioned nobody here on Q&A can read backend provisioning logs. Those are only visible to a support engineer with a case open. If you do open one, raise it as Technical, not a quota request, otherwise it gets routed to the team that handles capacity and they'll come back saying there's no capacity to grant.

Post the expanded error and we can take it further.