Serverless Application Engine

Serverless Application Engine (SAE) is the world's first application-oriented serverless PaaS, providing a cost-effective and highly efficient one-stop application hosting solution.

SAE Benefits

Out-of-the-Box Microservices and Kubernetes Clusters
The out-of-the-box microservice environment allows you to use Spring Cloud, Dubbo, and HSF applications with zero code modifications. WAR, JAR, or PHP packages can be deployed directly without containerization.
Low Cost and Ultimate Performance
Fully-managed IaaS with Kubernetes clusters provide low-cost solutions for your business. SAE scales within seconds and improves the efficiency of runtimes and Java application startup.
High Availability and Security
Network isolation with sandboxed containers and virtual private cloud (VPC) ensures the security of application runtimes. SAE provides high availability solutions for large-scale events that require precise capacity handling, high scalability, and service throttling and degradation.
One-Stop PaaS
One-Stop PaaS with seamlessly integrated basic services, microservices, and DevOps products

SAE Features

Full-Lifecycle Application Management

SAE provides full-lifecycle application management. You can implement different release policies, such as phased release and canary release. The traffic-ratio-based canary release model is also supported. The release process is fully observable and can be rolled back.

Microservice Management

Spring Cloud, Dubbo, and HSF microservice frameworks can be migrated to SAE with zero code modifications. SAE also provides capabilities, such as service registration and discovery, environment isolation, configuration management, service management, service throttling and degradation, seamless application publishing and unpublishing, and service authentication.

Automatic Creation of Docker Images

SAE creates Docker images for WAR, JAR, and PHP packages automatically. You can deploy non-containerized services directly on SAE.

Auto Scaling

You can set scaling policies with web-based counters (for example, trigger scaling by monitoring counters or at a scheduled time), improving the responsiveness of application scaling. SAE can scale within seconds to handle unexpected traffic surges during peak hours and reduce costs during off-peak hours.

One-Click Switch for Test Environment

Resources are wasted when application instances are retained in the idle test environment of large- and medium-sized enterprises during non-working hours. You can use SAE to enable or disable the test environment in one click, which can save up to two-thirds of operating costs for the test environment.

Comprehensive Monitoring

SAE provides the monitoring function for systems (CPU and memory), applications (QPS, RT, operation frequency, SQL status, and the number of errors), and lifecycle (trace status). You can identify and address issues using the monitoring function.

Accelerated Cold Start of Java Applications

Application Class-Data Sharing (AppCDS) using Dragonwell 11 accelerates the startup of Java applications by eliminating the need for initialization, class loading, and just-in-time (JIT) compilation. This speeds up the cold start of Java applications by 40%.

Remote Debugging and Interconnection between On-Premises and Cloud Services

It supports remote debugging for diagnosing Java applications deployed in SAE. You can also subscribe to on-premises services and register these services in the SAE registry. Operations can be called between on-premises and cloud microservices.

Microservices Comparison

Item

Open-Source Spring Cloud/Dubbo + IaaS Clusters Open-Source Spring Cloud/Dubbo + Self-Managed Kubernetes Clusters

SAE

IaaS Purchase & System Building

Yes Yes No. SAE features a built-in microservice framework and Kubernetes clusters.

O&M Costs

Yes Yes Fully-managed

Hardware Costs

The annual subscription of IaaS based on the peak traffic volume is relatively costly due to low overall resource utilization. The annual subscription of IaaS based on the peak traffic volume is relatively costly due to low overall resource utilization. Scalable resources and pay-as-you-go billing reduce costs by minimizing idle resources.

Application Management

Operations over command-line interfaces (CLIs) are inefficient and error-prone. Self-integrated cloud services and self-developed release policies are used to avoid upgrade downtime and service interruption. Operations on kubectl or Kubernetes Dashboard require a certain level of experience. SAE provides self-integrated cloud services and self-developed release policies to minimize upgrade downtime and service interruption. Web-based operations enable automatic traffic distribution based on the application status. SAE supports the canary release policy, and releases can be observed and rolled back. SAE can also be used to implement CI/CD with Jenkins, Yunxiao, Alibaba Cloud Container Registry (ACR), and source code.

Microservice Management

Integrated microservices are customized based on open-source frameworks. Integrated microservices are customized based on open-source frameworks. Microservice frameworks are migrated seamlessly to build a full set of stable microservice capabilities.

Permission Management

Self-developed solutions Permissions are granted with the role-based access control (RBAC) mechanism of Kubernetes, which does not meet enterprises' requirements of permission isolation among multiple tenants. SAE is fully compatible with Resource Access Management (RAM), providing permission isolation at the namespace and application levels.

Scaling Efficiency

Within minutes Within minutes Within seconds

Monitoring and Diagnosis

Users need to manually integrate diagnostic tools and maintenance servers based on the open source system and build alerting channels. Users need to manually integrate diagnostic tools and maintenance servers based on the open source system and build alerting channels. SAE provides the monitoring function on systems, applications, and traces. SAE also provides application diagnostic capabilities and supports customized alert rules.

Application High Availability

The development of high availability architecture requires a team of experts. The development of high availability architecture requires a team of experts. SAE supports multi-zone deployment and provides high availability solutions for large-scale events that require precise capacity handling, high elasticity, and service throttling and degradation.

User Experience

Users need to perform many steps (such as purchasing infrastructure, creating clusters, and initializing clusters) before creating applications. Interconnection to cloud services also requires manual operations, which results in an inferior integration experience. Users need to perform many steps (such as purchasing infrastructure, creating clusters, and initializing clusters) before creating applications. Interconnection to cloud services also requires manual operations, which results in an inferior integration experience. Applications can be created easily. Seamless interconnection with various cloud services delivers a one-stop PaaS experience.

SAE Scenarios

Easy-to-Use Microservice Framework Transformation, Accelerating Iteration Efficiency

In enterprises' production environments, microservice applications are separated and converted into WAR, JAR, or Docker images that are stored in ACR. You only need to iterate each microservice application based on Spring Cloud or Dubbo development standards. SAE provides the capabilities to deploy and schedule underlying resources, implement canary releases, manage microservices, and monitor and diagnose services.

Benefits

  • Smooth Application Migration with Zero Code Modifications

    Dubbo and Spring Cloud applications can be migrated to the cloud easily with zero code modifications.

  • Fully-Managed Infrastructure (IaaS, Kubernetes, Microservice Components, and APM Components)

    The infrastructure is fully-managed, so you do not need to build the ZooKeeper, Eureka, Consul, and Skywalking manually. This reduces the development and O&M costs significantly and ensures stability in commercial use.

  • A Full Set of Out-of-the-Box Beginner-Friendly Microservices

    SAE implements automatic image creation and provides advanced features for enterprises, such as canary releases, traffic control, environment isolation, application monitoring and diagnosis, and permission management. SAE is easy to use even for users that do not have much containerization experience.

Scalable High Availability Solutions for Large-Scale Events

Unexpected traffic surges are common in new retail, e-commerce, online education, and entertainment industries. SAE provides a complete set of high availability solutions with precise capacity handling, high scalability, and service throttling and degradation capabilities.

Benefits

  • SLA Guarantee in Peak Hours

    SAE auto scaling can be used together with service throttling and degradation provided by Application High Availability Service (AHAS) to help handle traffic surges and guarantee service reliability.

  • Fast Diagnosis and Fault Location

    SAE integrates seamlessly with Application Real-Time Monitoring Service (ARMS) to provide web-based application monitoring and diagnosis. This enables fault identification for slow SQL queries, slow operations, and stacks for called operations. ARMS facilitates the analysis, troubleshooting, early warning, and resolution of online issues, saving a lot of time.

  • Optimal Efficiency and Minimal Costs

    SAE provides various scaling policies, achieving scaling in seconds and ensuring minimal hardware operating costs. You do not need to pay for idle resources.

Multiple Deployment Modes for Optimal Continuous Delivery

Yunxiao, Jenkins, source code, Cloud Toolkit plug-in, or ACR can be used to deploy images and packages. SAE completes the DevOps process from code submission to application deployment automatically, which serves as an efficient substitute for the complicated and slowly-iterated traditional deployment methods.

Benefits

  • DevOps Automation

    DevOps automates the whole process of code modification, code building, image creation, and application deployment.

  • One-Click Deployment to the Cloud

    Alibaba Cloud Toolkit plug-in is used to deploy on-premises applications to the cloud with one click, facilitating remote debugging.