I am working on a project that passes Graph events through an Azure Event Hub to trigger a function app. The subscriptions had been set up and running for months without issue. Two days ago, I switched my functions over to the isolated model as opposed to the in process model. Before this switch over, subscriptions to Graph with the Event Hub passed into it as notificationUrl in the form below was working.
EventHub:https://{eventHubNamespace}.servicebus.windows.net/eventhubname/{eventHubName}?tenantId={tenantId}
While swapping the function apps over, I received a notification there was an issue with Host ID overlaps. After resolving this error, I went to add a new subscription to Graph before I received this error, which I had not encountered to this point:
{
"error": {
"code": "ValidationError",
"message": "Subscription validation request timed out.",
"innerError": {
"date": "2025-06-20T20:35:36",
"request-id": "xxxxxxx",
"client-request-id": "xxxxxxx"
}
}
}
I attempted to resolve this issue by adding a new event hub to the same event hub namespace I had been using, but that did not solve the time out issue. I have also created a new namespace, and tried with the event hub in my other resource group.
The event hubs are still actively processing current Graph Subscription notifications, but my re-subscribe function that maintains these subscriptions is obviously failing with the same issue. It isn't a problem with the bearer token either, as I am using the same method as always. The only other change I can think of that would impact this would be some sort of throttling but there are less than a hundred active subscriptions to graph and less than a few thousand events every day. There are currently as of posting time no listed outages for Azure, but I did find that earlier this month there were some Azure outages reported. I am within a tenant set up in West US, all resources are set up in West US, and I am using an app registration for authentication into Azure. These requests were made both through C# code and through Postman, and attempted on multiple machines, all with the same response of "timing out". My event hubs are enabled and active, showing no errors or raising no warnings. Any help would be great. Thanks.