Exclusive Dialogue with Volcano Engine Zhang Xin: The next 3 years will usher in the peak period of cloud native containers


Exclusive Dialogue with Volcano Engine Zhang Xin: The next 3 years will usher in the peak period of cloud native containers

There is no uniqueness in the choice of technology paths by enterprises. Under the current wave of strong digitalization, cloud native has become a popular direction of technology. The combination of virtualization, containers, and microservices to provide flexible enough cloud services for enterprise business will become the next universal technology stack.

Beginning with the deployment of containers, ByteDance’s cloud-native practice has gone faster and faster in recent years.

In June 2016, ByteDance enabled the Kubernetes technology stack and began to carry out large-scale containerization transformation of the business. In 2017 and 2018, it entered the stage of mass production. By 2018, the number of single clusters of containers deployed in this year has reached tens of thousands. node.

Today, more than 95% of ByteDance’s applications have achieved cloud nativeization. From the initial container-oriented deployment to the service-oriented stage, not only the container orchestration tool is used, but also a comparison is established on this. Perfect microservice governance system. In 2020, Volcano Engine, an enterprise-level technical service platform under ByteDance, officially opened up its cloud-native capabilities.

In addition to not having too much historical baggage, this young company can go into battle with light luggage without paying attention to the cost of replacement, but also has more judgment and investment in technology trends, and has embraced the cloud approach very early. Therefore, many of ByteDance’s businesses are deeply ingrained in the cloud from the beginning, and when designing new business systems next, they will also fully consider the elastic capabilities of the cloud. Born and bred in the cloud, this has become the most straightforward and universal cloud-native approach that enterprises can learn from.

Zhang Xin, deputy general manager of Volcano Engine, explained his judgment on the trend of cloud-native container technology, the difficulties and challenges faced in introducing this technology, and ByteDance’s own cloud-native practice path in the past five years.

“In terms of the technical curve, virtualization has experienced almost ten years of development, and now it has reached a stable period of commercialization. Similarly, container + K8s has been open sourced by Google in 2014 to the present, and it is expected to reach a peak in the next three years. .”

However, due to the wide range of dimensions involved in Kubernetes, including the concept of microservice architecture, network, storage and other levels, both product service providers and enterprise practitioners have no small technical challenges in this process.

Zhang Xin, as the head of the cloud native application service of Volcano Engine, is also an early researcher and practitioner of the domestic container industry. A full summary of the interview is presented below.

Get to know the four directions of cloud native

Although cloud native is listed as the main development direction by major internet companies at the technical strategy level, cloud native and containers are two related concepts that cannot be equated.

At the beginning, what the outside world may be more concrete and easy to understand is the container itself, which is a standardized encapsulation of resources and applications. However, having a containerized deployment may only mean that the enterprise has a cloud-native infrastructure, but it does not mean that cloud-native applications have been implemented. The evolution from container to cloud native will still be quite different.

Judging from the perspective of cloud native core technologies, there should be four evolutions in the four directions of southeast, northwest and northwest when cloud native enters the mature stage:

  • Let’s look south first, facing the new generation of cloud-native infrastructure. Infrastructure as code (IaC), as a popular technology trend in the recent period, aims to make IT infrastructure more efficient and agile, deploy the infrastructure faster in a configurable way, and further improve the overall architecture. Performance, Agility. Domestic cloud manufacturers have also launched many similar products or technical solutions.
  • Look east and west again. There are two trends: one is that more applications and middleware, including big data platforms, databases, data warehouses, AI platforms and other emerging business loads can be cloud-native or even serverless; It can effectively realize the opening of data silos in the enterprise and the connection of systems, including API services, API governance, and API gateways.
  • Then northbound. In this direction, a number of emerging products for agile development have emerged, such as low-code and no-code tools, as well as development and testing scenarios, Industrial Internet and other industry scenarios, as well as BPM for business scenarios, etc., which can be based on cloud-native microservices technology to build aPaaS for application development.

In terms of technology curve, virtualization has experienced almost ten years of development, and now it has reached a stable period of commercialization. Similarly, containers (the technical standard announced by Docker in 2013) + Kubernetes (Google’s open source project in 2014) are expected to reach a peak in the next three years. Judging from the status quo of industrial digital development, all walks of life are now in full swing, and industry digitalization is accelerating. In theory, it is a very good window period.

A cloud native user survey report by the China Academy of Information and Communications Technology shows that containerization is still the mainstream trend, and the number of respondents who put container technology into core business production increased by 2.48% year-on-year. According to data released by research firm Gartner, by 2022, 75% of global enterprises will use cloud-native containerized applications in production.

But there is also a huge gap in the middle, that is, how can the above-mentioned technical products in the east, west, north and south directions be closely connected with the front-end business needs. This is something that companies, including suppliers, need to quickly implement in the next three years.

top priority task

So, what will the volcanic engine first land on these four levels?

“What we are doing more is the north-south direction. That is, we cultivate our internal strength downwards and provide services for developers upwards. The southward direction is the most critical internal strength in our opinion. Any cloud application or system must eventually return to stability. This is something we must do and are doing very well at present.” Zhang Xin pointed out.

First of all, the byte-beating business itself is faced with the problem of how to maintain a sufficiently good performance in the case of large-scale data volume and business volume. ByteDance has a very large data center that houses nearly one million servers. At the cloud native infrastructure level, including how to better combine computing, network, and storage with container technology, the team has made a lot of technological innovations.

Second, cloud-native applications must eventually be provided to developers. How to make development faster and easier is an important priority pursued by the team.

However, the current enterprise practice of containers can also be divided into three stages:

  • In the first stage, for container orchestration management, users directly deal with container Kubernetes;
  • In the second stage, further improvement began, and users changed from container-oriented to service-oriented, so there was governance of various microservice frameworks. At this time, users did not need to pay attention to container configuration, management and other issues, but to consider How to split the entire business architecture, what microservice components should be, and how to interconnect and interoperate between different components;
  • The third stage is the serverless stage. At this time, users do not need to care about how to design the overall business architecture, but only need to pay attention to how the front-end business logic can be expressed quickly, which also requires manufacturers to advance the middle and background components behind the business. preset.

It is worth mentioning that ByteDance itself has a middle-office organizational form. It will support many business lines to build internally, and you can see many applications or functions. In fact, it is built based on a serverless approach.

At present, most enterprises are still in the first and second stages of container-oriented and service-oriented, and at the product level, ByteDance has begun to focus on the third stage.

Exclusive Dialogue with Volcano Engine Zhang Xin: The next 3 years will usher in the peak period of cloud native containers

an emerging force

At the customer level, different types of customers will have obvious differences in introducing cloud-native container services.

Among them, the Internet industry has become the industry with the highest proportion of cloud native adoption. First of all, Internet companies have a great depth and breadth of cloud-native applications, and are willing to pay for high-ROI technologies.

Especially the leading Internet companies, which have extremely strong resource capabilities and complex business forms, often choose the method of open source in-depth customization + self-research for technical practice, because their business volume is large enough to pass even 1% of the business. Improved performance and higher ROI.

In contrast, it may be difficult for many Internet companies with waists to invest a lot of R&D teams in order to improve performance or cost by 1%. A recent survey by ByteDance also found that the customer base of the Internet industry is further divided, and the adoption rate of cloud native in vertical industries such as online games has exceeded 80%.

For traditional industries, the business-driven impact of digital transformation has made them increasingly embracing new technologies in recent years.

Industries such as finance, retail e-commerce, industrial Internet, Shengxin medical care, and smart government, driven by digital business innovation, have more and more Internet-related businesses. From the team’s recent service observations, these industries are increasingly embracing new technologies.

For example, in the field of Shengxin Medical, with the continuous emergence of businesses such as protein sequencing and gene analysis, and the advancement of the digital construction of medical infrastructure, enterprises need more and more computing power to build more agile application systems. In addition, the Shengxin medical industry itself is also striving to build a basic scientific research platform to open up the problem of data silos where hospitals and universities have repeated wheel-building in scientific research.

Do businesses need to pursue new technologies?

It is said that Google has implemented containerized applications for itself as early as 2003, and it has been kept secret as a secret weapon for many years. Relying on this technology, it can save half of the cost of operation and maintenance every year, which undoubtedly greatly enhances the company’s own competitive advantage.

Zhang Xin, who used to work in this company, has witnessed Google’s 2 billion container computing clusters running every day.

In his view, for enterprises, they will not simply pursue the pursuit of new technologies. Trying new technologies still needs to be driven by business. Therefore, if you divide it from a business perspective, you must first see whether the enterprise has enough agile businesses, and then further measure the ROI of using new technologies for transformation based on this.

The extension of container technology in more stateful application scenarios has led to the continuous optimization of the IT infrastructure, such as the capacity requirements for storage and databases, the complexity of the development and operation environment, and the need for large-scale Calculate the existing shortcomings.

From the application point of view, stateless applications, stateful applications, steady-state applications, etc., the ROI and difficulty of cloud nativeization of different applications are different.

  • First, stateless applications are most suitable for the first batch of cloud-native transformations.
  • Second, stateful applications are often clustered applications, including business scenarios such as middleware, databases, and Elasticsearch.To solve this part of the scenario problem requires certain technical barriers, but once the stateful application is cloud-native, there will actually be better returns.

In the container environment, a problem that stateful applications often encounter is how to persist data. This has changed the requirements for enterprise storage. For example, which storage system to choose, can it meet the performance of distributed storage? Can the data be combined with the upper-layer application orchestration? Do you need to consider the affinity of data when scheduling services?

After that, distributed storage itself also depends on a strong network system, which requires the container itself to have its own corresponding solution. How to perform high-performance transmission based on RDMA, whether it can be directly at the hardware level, such as hardware offloading based on smart network cards… and so on these technical difficulties.

But from another perspective, these problems also provide a lot of room for many cloud-native container vendors. Because these capabilities cannot be solved within the open source technical system, this is also the area where Volcano Engine hopes to exert its strength.

  • The third category is traditional or steady-state applications.

For such applications, it will be difficult to migrate first, they may be some commercial package software, and they may run on non-x86 architecture; secondly, if such applications are built, they may not generate a lot of returns There may be no need to address performance issues such as high elasticity, no need for agile iterations, and it’s not even recommended to spend a lot of effort and risk in exchange for rewards.

New Opportunities in the Cloud-Native Era

Cloud native will also affect the understanding and consensus of various cloud vendors on standards.

One is to form standards. Otherwise, there is a closed system between different manufacturers, and it will be difficult to be compatible with each other, and it will be difficult for the manufacturers themselves to become bigger, and the enterprises themselves will also face more and more technology silos.

Second, cloud native will drive the formation of standards. Previously, the popularity of container Docker was because it defined a set of management standards for application packaging, packaging, and operation. Kubernetes also established a set of standards for how containers are orchestrated. Although there will be more and more technologies such as serverless, aPaaS, low-code, etc. in the future, containers + Kubernetes have actually solved many of these fundamental standardization problems.

“This state is relatively optimistic.” Zhang Xin finally pointed out.

Differences in understanding of openness vary due to the balancing strategies of commercial interests involved in different standards bodies. In the current market, the competition situation is still in the hands of the economies of scale of international cloud vendors. How can we make breakthroughs at the technical level? In a winner-takes-all market, working together on open standards and developing open source software together to offset the dominance of the market leader may be the only way to go.

The Links:   3HAC023518-001 3BSE008534R1 DISPLAY


Pre:    Next: