This topic describes the instance specifications of the original API Gateway. API Gateway instances are classified into dedicated instances and serverless instances. In addition, API Gateway provides the virtual private cloud (VPC) integration instance and dedicated instance cluster to meet specific business requirements.
API Gateway instance types
An API Gateway instance is a set of resources that are used to access your APIs and process requests to your APIs. API Gateway instances are available in the following types:
Dedicated instances (VPC): The technical specifications of dedicated instances (VPC) and shared instances (VPC) are the same. To obtain a higher service-level agreement (SLA) guarantee, you can purchase a higher specification for requests per second (RPS) to use dedicated resources. The resources include inbound public IP addresses, inbound IP addresses in VPCs, outbound public bandwidth, and isolated server clusters.
Serverless instance: Serverless instances are previously named shared instances. Shared instances are upgraded to serverless instances region by region. Workloads running on shared instances are not affected by the upgrade. Serverless instances are billed based on the same method and prices as shared instances. Serverless instances provide higher scalability and performance than shared instances and are more suitable for scenarios with a changing traffic volume and that require timely O&M-free scaling.
For more information, see Choose an instance type.
Change an instance specification
If a dedicated instance does not meet your business requirements, you can upgrade or downgrade the instance specifications in real time.
You can change specifications only for instances that are running.
If you upgrade the specification of a subscription instance, you must pay the price difference between the existing specification and the new specification. If you downgrade the specification of a subscription instance, you will be refunded the price difference by using the original payment method. Redeemed vouchers are not returned.
The new specification takes effect immediately after the specification change is complete.
The specification change takes about 5 to 10 minutes to complete. During the specification change, APIs that are running on the instance can still be called.
For more information, see Change an instance specification.
VPC integration instances
API Gateway provides the VPC integration instance to facilitate the communication with your VPC. The instance is suitable for the following scenarios:
API Gateway needs to access multiple resources, such as Elastic Compute Service (ECS) instances and Server Load Balancer (SLB) instances, in the same VPC. These resources are the backend services of an API hosted on API Gateway. Typical scenarios include microservices and service discovery, such as Nacos.
Hybrid cloud network communication is used. In this case, a backend service accepts only the private IP addresses of Alibaba Cloud VPCs in the requests forwarded by API Gateway instances. For more information, see the Scenario 3: Access a backend service deployed in an on-premises data center from an ECS instance deployed on Alibaba Cloud section of the "Centralized API management on a hybrid cloud" topic.
For more information, see VPC integration instances.
Dedicated instance clusters
API Gateway allows you to cluster dedicated instances to utilize their combined performance. If a single dedicated instance of the largest specification cannot meet your business requirements, you can cluster multiple dedicated instances of the same type and specification in the same region to combine their capabilities to provide services. This helps meet your performance and scalability requirements in heavy-load scenarios by exceeding the limit of individual instances. For more information, see Dedicated instance clusters.