Serverless Infra Architecture: What Benefits It Gives and How Will It Change in 2020
This is the sixth post in Seven Major Cloud Native Trends for 2020 Series
The evolution of Kubernetes-based serverless infra architecture has always been the focus of major cloud vendors and communities. In December 2019, AWS announced the official GA of EKS on AWS Fargate at the annual re:Invent conference in Las Vegas, which caused considerable excitement in cloud markets and communities.
EKS on Fargate provides a standard serverless infra user experience. After purchasing the EKS service, you can use native Kubernetes APIs to deploy your applications without having to purchase additional infrastructure cloud resources (such as VMs and Nitro). In addition, pay-as-you-go is supported.
With a serverless infra architecture, you do not need to pay attention to the details of the underlying infrastructure such as computing, network, and storage. This means you can truly return to the pod-oriented application resource deployment pattern. The strong isolation between the control plane and the data plane will be the key value of serverless infra architecture. In addition to masking users from the details of the underlying infrastructure, serverless infra must provide a secure and trusted tenant isolation environment.
As the Alibaba economy was fully migrated to the cloud in 2019, the underlying scheduling system was upgraded to the cloud native Kubernetes + lightweight container architecture and deployed on a large number of ECS bare metal instances. At the same time, Kata Containers-based security container runtime technology is maturing and has already been widely deployed.
As we enter the new decade, we expect that 2020 will be the year that the economy fully moves toward serverless infra. The serverless infra architecture will be based on the X-Dragon + security container architecture. By building software and hardware multi-tenant capabilities, elasticity capabilities, and powerful container self-recovery capabilities, the architecture will provide users with an extremely secure, stable, and isolated user experience. At the same time, the sharing of the underlying resource pool can effectively improve overall resource utilization, and resource intercommunication after the pool sharing effectively reduces overall machine costs.
You May Also Like
This article narrates the evolution of container technology over a decade and how it reforms the modern software infrastructure and delivery.
This article discusses some of the challenges Alibaba faced and had to overcome in implementing its own service mesh solution for the big annual Double 11 shopping event.