How to prevent interruptions of gateway availability during API management service deployments?

Question

We deploy our API management service (sku premium tier, virtual network type internal) via bicep and observe during deployments that the gateway is not available for some time (usually 15-20min). In the deployments overview of the affected resource group, we notice two operations of approx. the same time:

• Provisioning operation Read of type Microsoft.ApiManagement/service
• Provisioning operation Create Microsoft.ApiManagement/service

Interestingly, the non-availability occurs even if no changes to the API management service would be made as a az deploy what-if shows.

Apart from the possibility to rework our bicep code to control deployments specifically of the API management service, which other possibilities do we have to avoid gateway non-availability during deployments?

Answer 1

Hello , Welcome to MS Q&A

If I understood right , I think to address the issue of the API Management service gateway being unavailable during deployments, you can consider implementing a multi-region deployment. This approach allows you to add regional API gateways to your existing API Management instance, which helps to reduce request latency and improve service availability.

In case a region goes offline, API requests can be automatically routed to the next closest gateway, ensuring that the API Management service remains operational. Additionally, during deployments, you can temporarily disable routing to a specific regional gateway by updating the gateway's disableGateway property to true, allowing you to configure and test the backend service without affecting the overall availability.

References:

• Deploy an Azure API Management instance to multiple Azure regions
• Ensure API Management availability and reliability

Please accept answer if it helps

Thanks
Deepanshu