The future has come: the development trend of serverless from the perspective of ICT serverless standard.

Serverless architecture (Serverless) is an architectural concept that completely separates applications from infrastructure. Developers don't need to care about the operation and maintenance of infrastructure, but only need to focus on the development of application logic, which truly realizes flexible scaling and on-demand payment. At present, the internal service serverless upgrade of major cloud service providers and head Internet enterprises has begun to test the water in a small scale; Server-less SME business applications have also begun to take shape, and enterprise-level applications have begun to take shape, and the future is expected. The serverless ecosystem has begun to take shape, and it can be predicted that serverless will become the trend of the next generation of cloud computing services.

Under this background, Cloud Function (SCF), Elastic Microservice (TEM) and Elastic Container Service (EKS) were combined with other related products, and * * * obtained the first batch of advanced certifications with the highest technical capability of serverless platform in 202 1 China.

In July this year, at the "202 1 Trusted Cloud Conference" jointly sponsored by China Information and Communication Research Institute and china communications standards association, Tencent Cloud won five awards and 10 trusted cloud certifications, and obtained 54 trusted cloud certifications in the evaluation of cloud storage, serverless and other sub-sectors, ranking first among domestic cloud vendors. Tencent Cloud Function (SCF), Elastic Microservice (TEM) and Elastic Container Service (EKS) have deeply participated in the formulation and implementation of serverless standards, and the platform technical capabilities provided by Tencent Cloud Serverless Product Matrix have also been unanimously recognized by peers.

Through this serverless standard, we will bring you the following interpretation of the development trend of serverless:

When we combine the concept of Servless with these products, Servless file system (CFS), database (TDSQL-C), gateway (API Gatgeway) and middleware (TDMQ) can greatly reduce the development and operation costs of Servless applications, let developers really focus on the core competence of business, and maximize the core R&D strength and IT investment of enterprises. Through the final demand-driven, we can foresee that serverless cloud service products may be the only way for the future development of cloud computing.

In the past, scene-based FaaS was the mainstream application form of serverless, and the landing cases were mostly lightweight sites, SSR and "adhesives on the cloud". In the process of formulating the serverless standard, there are rich discussions and exchanges on how to evaluate the actual serverless landing form of enterprises. We think that the application form of serverless can be FaaS, microservice or even a single application; The runtime environment can be a native runtime or a container image; When landing, it can be used to provide API interface, run audio and video transcoding, live streaming and other computing tasks, and complete site pressure measurement, AI reasoning and other tasks.

However, it is impossible to make the existing stock system serverless overnight, which is a process of continuous design and revision. Serverless applications also need to go through several stages, such as migration, optimization and cloud native architecture transformation, and there needs to be a smooth switching process between different stages. With the help of the Web function of cloud function, the migration process can be smoother. Only when the actual load is running on the server-less, can the optimization and cloud native transformation of the server-less application be continuously carried out based on the actual running results of the production environment and the collected quantitative indicators, so as to further exert the value of the server-less.

When the services for building applications are gradually migrated to the cloud, the development environment is further "cloud-oriented". Compared with local development, it also faces some new challenges, such as code effective time, local testing, remote debugging and offline development. These are the key links that affect the efficiency of developers. In this standard of "technical capability of serverless platform", the support of tool chain is regarded as one of the important dimensions to measure the technical capability of serverless platform. A mature serverless developer platform needs to provide friendly ide support, so that developers can use familiar development tools to develop serverless applications and reduce the switching cost of developers. In addition, in local or remote testing, good tool support is needed, which can easily initiate calls, trigger application execution and quickly return results. When the results do not meet expectations, a series of troubleshooting methods such as monitoring and logging are needed to help developers quickly locate the problem.

As the most popular one-stop development tool in the serverless community, the serverless framework has millions of active applications and 50,000+daily downloads. As early as 20 19, we cooperated exclusively with Tencent in Greater China without a server framework, and deeply integrated with serverless products such as Tencent Cloud's cloud function. At the same time, the community also has a large number of plug-ins and out-of-the-box templates to help developers get started with serverless application development quickly. In addition, cloud development is also the largest WeChat applet application development platform in China. Small program applications such as Sichuan Tianfu Health Link and Shenzhen Airport Smart Travel Service all run on Tencent Cloud's serverless platform.

Server-less Cloud Function (SCF) is a server-less execution environment provided by Tencent Cloud for enterprises and developers, which helps you run code without purchasing and managing servers. Just write the core code and set the conditions for the code to run, and you can run the code flexibly and safely on Tencent Cloud infrastructure.

By simply modifying the listening port, the popular Node.js framework can be directly deployed to the cloud, enjoying the advantages of no operation and maintenance, low cost and on-demand expansion and contraction brought by serverless technology.

Breaking through the execution time limit of traditional FaaS products, the first FaaS product supports long-term tasks of up to 24 hours, and supports large-scale audio and video file processing, live streaming, data analysis and other scenarios.

It supports the allocation of120 GB (122,880 880MB) large memory environment for the first time in the industry, and can more easily handle workloads with higher memory or more intensive computing requirements, such as audio and video processing and big data analysis.

Smooth migration of applications to cloud functions through Web functions and containerized mirroring supports hosting H5 pages, APIs, SSR applications, applets and other forms of applications, shortening the research and development cycle, quickly collecting market feedback and accelerating product iteration.

There is no need to operate and maintain virtual machines or other computing clusters. With the extreme flexibility and quantity charging provided by cloud function, audio and video recording, transcoding, mixed streaming, editing and streaming can be carried out efficiently and at low cost, so that enterprises can focus on the audio and video processing logic itself, thus continuously improving the content quality and optimizing the audio-visual experience.

You can connect with other cloud services through triggers, such as object storage (COS), logging service (CLS) and other services. When the upstream data transmission changes, it will automatically trigger the function to execute the calculation logic. Typical usage scenarios include: CDN refresh and preheating, middleware messaging, file backup, etc.

It supports various trigger functions, such as timing and message queuing, to execute output processing logic and perform data processing operations, such as data collection, data cleaning and ETL. The processed data can be directly stored in the downstream data warehouse, business database or BI analysis system.

Tencent Cloud Elastic Microservice (TEM) is a serverless PaaS platform for microservice applications, which realizes the perfect combination of serverless and microservice, and provides out-of-the-box microservice application hosting service with zero application transformation, quantity payment and no operation and maintenance.

Flexible microservices embrace open source, support microservice applications such as Spring Cloud to the cloud, and provide capabilities such as application running hosting, service registration discovery, microservice governance, and multi-dimensional monitoring. To meet the needs of consulates, Eureka and other registration centers. Flexible microservices help you create and manage cloud resources and provide second-level flexibility. You can use it on demand and pay by quantity, which greatly reduces resources and operation and maintenance costs, so that you can fully concentrate on the core business logic of the enterprise and help your business succeed.

Elastic microservice provides customers with out-of-the-box microservice solutions through application hosting, service registration and discovery, service governance, call chain and multi-dimensional monitoring. Help enterprise users quickly build micro-service applications, greatly improve operation and maintenance efficiency, reduce the complexity and technical threshold of service governance, let enterprises focus on core business and help customers succeed.

In the scene of tidal characteristics and sudden traffic, it is easy to have problems such as overtime access response and increased bit error rate. Tencent Cloud Elastic Micro-service provides second-level elastic scalability to help enterprise customers easily cope with traffic peaks.

Tencent Cloud Flexible Micro-service helps customers to continuously integrate delivery and realize rapid iteration of micro-service applications. From code development to application delivery, elastic microservices provide the ability of ide plug-ins, gray publishing and other publishing strategies to help corporate customers quickly verify business value.

Eks (elastic Kubernetes service) is a serverless kubernetes service launched by Tencent Cloud Container Team, which can directly deploy workloads without users buying nodes. It is fully compatible with native Kubernetes, supports native purchase and management of resources, and charges according to the actual amount of resources used by the container.

Whether it is a self-built K8s cluster or a cluster hosted by Tencent Cloud TKE, as long as the network is interconnected, it can expand the cluster resource pool by deploying EKS virtual nodes at almost no cost. When expanding Pod, Pod can be quickly dispatched to Tencent Cloud public cloud resources corresponding to "virtual nodes" automatically or manually.

Compared with the traditional scheduling of resources by expanding and contracting servers (heavy process and long time consuming), virtual nodes provide an ability to directly schedule Pod, which can be faster and more efficient.

Elastic container service EKS is used to run micro-services, which avoids the operation and maintenance of computing nodes by users. Services can automatically scale according to the load, use the most reasonable amount of resources to carry applications, and reduce the cost of resource use.

Using the elastic container service EKS to run offline computing tasks, you can quickly deploy the task load by preparing the container image. In addition, the elastic container service EKS only charges the computing power used in the actual operation of the task, and the Pod is automatically released when the task ends, and the charging ends.

The elastic container service EKS supports the use of CPU, GPU and vGPU to run online reasoning services. Rich resource specifications and flexible loads make running services more efficient and economical.

Experience the server-less demonstration of Tencent Cloud immediately and get the new user package of server-less Tencent Cloud server-less novice experience.