Forum Discussion
Downtime of API Management during deployment
Which actions will cause a downtime of Azure API Management so APIs will not be able to consume (Premium Tier)?
E.g. changing network would be obviously. But what about deployments of APIs, Products, Scaling up and down...
How can I find out which actions will make the APIs unavailable? Or is there a mechanism which will keep them available in each scenario (Even if "Service is being updated")?
Thanks!
5 Replies
For Premium (and the other production tiers), most control-plane operations are rolled out in a way that keeps the gateway processing traffic. The “Upgrade and scale an Azure API Management instance” article explicitly states that infrastructure changes such as scaling, custom domains, vNet config, etc. are applied as rolling updates and the gateway keeps serving requests, except for the Developer tier. Microsoft Learn
In practice, what we plan for is:- Short windows with reduced capacity while units are updated, based on the capacity metric guidance (watch for sustained values above ~60–70% before you change units or tier). Microsoft Learn
- Possible brief impact when doing bigger changes like vNet integration or region adds, so we schedule those in low-traffic periods. Microsoft Learn+1
If you need an even tighter availability posture, you can combine Premium with multi-region deployment and a traffic manager / Front Door layer in front of APIM, so you can drain one region while the other keeps serving traffic. Microsoft Learn+1
In the Azure API Management Premium tier, the majority of configuration activities such as deploying APIs, updating products, or scaling out are performed without service disruption. However, infrastructure-level modifications, including changes to virtual network integration, IP configurations, regional migrations, or pricing tier adjustments, may result in temporary service interruptions. While architecture is designed to maintain API availability during routine updates, operations that alter the underlying network or service tier can cause APIs to become temporarily inaccessible.
shailesh14Microsoft
You will not have downtime in premium tier, the SLA is 99.99% however even in standard tier there will be more than 1 node to serve your API request hence you may not observe downtime.
However, while service is in upgrade or update mode some operations like updating certificates, changing schemas may be locked in the period of time- This can be done via rules/triggers that can be implemented leveraging ARM, Bicep, Visual Studio Code, and PowerShell. An ARM is an IaaS (Infrastructure as a Code) template is a JavaScript Object Notation (JSON) file that defines the infrastructure and configuration for your APIM. The template uses declarative syntax. This will give you the flexibility to manage your APIM including other Azure resources for creating, updating, and deleting them.
