Custom domain stuck in "Deleting" across multiple Static Web Apps — blocks deletion and re-registration

[Vaishakhi]

Describe the bug

A custom domain (an apex domain and its www subdomain) is stuck in status
"Deleting" simultaneously on several Static Web Apps in the same subscription.
The portal shows: "An unknown error has occurred while adding your custom domain.
Please try again later."

Because SWA custom domains are globally unique, the hostname is now treated as
in-use: it cannot be added/validated on ANY app, and every app still holding it
fails to delete with Code: ResourceDeletionFailed. This is a deadlock only the
product team can clear.

Region / provider

• Region: East Asia
• Provider: Microsoft.Web/staticSites
• (Subscription ID, exact resource names, and correlation IDs - privately
  on request — happy to share via a secure channel.)

Steps that lead to it

1. Add a custom domain to a Static Web App before the DNS CNAME had fully propagated.
2. Validation fails; try to delete the custom domain to re-add it.
3. The delete hangs indefinitely in "Deleting".
4. Repeating on new apps spreads the same stuck hostname across multiple apps.

What I've already tried (all fail)

• Portal "Delete" on the custom domain → hangs in "Deleting" for >1 hour.
• az staticwebapp hostname delete → "Operation returned an invalid status 'OK'", entry remains.
• az resource delete --ids <.../customDomains/<host>> → same result, entry remains.
• az staticwebapp delete on the app → ResourceDeletionFailed (blocked by the stuck child domain).
• Confirmed there are NO management locks on the resource group.

Request

Please force-clear the stuck custom-domain bindings server-side so the hostname is
released, and/or unblock deletion of the affected apps. I can provide the
subscription ID, resource IDs, and correlation IDs privately.

Environment

• Deploying a Next.js static export via GitHub Actions to Static Web Apps (Free plan).
• CLI: latest az.

[jbrassell-boop]

Independent repro of what looks like the same partition-7 incident, different subscription, today (2026-08-12).

Symptom (identical): adding a custom domain sits at RetrievingValidationToken ~10 min, then fails — "An unknown error has occurred while adding your custom domain. Please try again later." (status Failed, validationToken null). validate/action succeeds, so CNAME/DNS is fine. A subsequent delete is Accepted, runs ~10 min, terminally fails, and the record is then stuck in Deleting indefinitely; az resource delete / raw ARM DELETE on the customDomains child don't clear it.

Reproduced across two apps + a control hostname: App A (East US 2) with both catalog.<redacted> and a throwaway swatest.<redacted> — both fail identically, ruling out anything domain-specific. App B, freshly created with --location "West US 2" — same failure. Both apps' default hostnames are on the .7. partition (...7.azurestaticapps.net, via azurestaticapps7.trafficmanager.net -> msha-slice-7-eus2-*).

New data point: I then created six brand-new SWAs across Central US, West Europe, East Asia, West US 2, and East US 2 — every one landed on partition .7.. No region/SKU/name produced a different partition. So all new SWAs in this subscription are currently pinned to azurestaticapps7, exactly the partition where custom-domain add/delete is failing — the incident appears scoped to that partition, and re-rolling can't escape it.

Correlation IDs and subscription/resource IDs available for backend lookup. Same-day siblings: #1794, #1796. Requesting (1) restore the partition-7 custom-domain workflow and (2) purge the stuck Deleting records server-side.

[nicoleannbustamante]

Hitting the exact same issue on a different resource (West Europe, not East Asia), so this doesn't appear to be region-specific.

Resource:
/subscriptions/8746eb58-7073-4ad7-9278-10796a6350f1/resourceGroups/rg-mavank-prod/providers/Microsoft.Web/staticSites/mavank-website/customDomains/www.mavank.com

Same root cause pattern: custom domain was added successfully via ARM/Bicep, later showed "Validation failed" with the same "An unknown error has occurred while adding your custom domain. Please try again later." message, and deletion attempts have been stuck in "Deleting" for many hours since with no completion.

One additional data point that might help narrow this down: the two control-plane views disagree on whether the resource still exists.

• az resource show --resource-type "Microsoft.Web/staticSites/customDomains" --name "<site>/www.mavank.com" (generic Microsoft.Resources provider) → 404 Not Found
• az staticwebapp hostname set --name <site> --hostname www.mavank.com (Static Web Apps-specific API) → returns the original resource record unchanged, same createdOn timestamp as when it was first created, still "status": "Deleting", same stale errorMessage

So the generic ARM RP thinks it's gone, but the Static Web Apps control plane still has it wedged. A direct az resource delete --ids <resource ID> against it also completed without visibly changing this state.

Same as the original report: this is blocking re-adding the domain entirely, and there's no self-service way out that we've found. Would appreciate any update on whether this needs manual backend cleanup.