Home Technology peripherals AI Volcano Engine elastic container instance: a powerful tool for transforming from a node-centric architecture to a serverless architecture

Volcano Engine elastic container instance: a powerful tool for transforming from a node-centric architecture to a serverless architecture

Mar 01, 2024 pm 08:04 PM
industry Lazy loading

As Kubernetes and container technology become increasingly mature, more and more enterprises are beginning to migrate their business applications to the cloud. By adopting cloud-native architecture, these enterprises can better support the rapid development and stable operation of their businesses.

As cloud computing develops in-depth into the development process, the maturity of Serverless architecture has brought cloud native technology to a new stage - Serverless takes the public cloud's elastic scalability, operation-free maintenance, fast access and other features to the extreme. It greatly reduces the user's cost of use, allowing users and enterprises to only focus on their business logic and achieve agile development in the true sense.

In order to better support enterprises in realizing the new serverless cloud native concept in business applications, the Volcano Engine cloud native team is comprehensively upgrading and innovating products, covering many aspects such as concept, system design and architecture design. . They are committed to giving full play to the advantages and value of Serverless technology to help enterprises run business applications more efficiently.

From Node Center to Serverless Architecture

The traditional Kubernetes architecture usually takes the node as the core, which means that the team It is necessary to build clusters on resource nodes such as cloud servers, and perform operation and maintenance management around these nodes. The traditional Kubernetes cluster architecture requires managing the deployment, monitoring, and maintenance of container applications on different nodes. This approach requires more manpower and time investment to maintain the stability and reliability of the entire cluster. However, with the development of technology, a new type of Kubernetes architecture has emerged, namely an architecture based on service mesh. In an architecture based on a service grid, network functions are abstracted and unified through specialized service grid components. As the business scale expands and the number of nodes increases, this architecture becomes increasingly important in container applications. There are usually some challenges in deployment and operation, mainly including the following aspects:

    Complex node operation and maintenance
  • : Users need to manage and maintain nodes by themselves, including node configuration. Complex and tedious node operation and maintenance such as initialization, operating system update and upgrade, security management, operation monitoring and log collection and analysis, fault recovery, network management, storage management, expansion and contraction, resource planning and allocation, and resource cost management Work;
  • Low resource utilization
  • : Nodes reserve fixed resources. Even if the actual workload requires less resources, the node resource usage cannot be reduced; the node resource fragmentation rate is high, Some unused resources distributed on various nodes are difficult to effectively aggregate and cannot be fully utilized; resources are over-allocated. In order to avoid performance bottlenecks, resources are often planned according to the maximum load, which makes it difficult to predict actual demand and may cause a large number of idle resources;
  • Slow resource expansion
  • : Container applications may experience bursts of traffic and need to scale quickly in response to increased load. In traditional node-centric architecture, the expansion of computing resources often requires the creation and configuration of new cloud server nodes. This process may take several minutes and cannot achieve instant expansion.
  • Volcage Engine
Elastic Container Instance

(Volcengine Container Instance, referred to as VCI) is a new cloud native solution launched for the above situation. It is abstracted from ByteDance’s internal cloud native technology for many years. Accumulation is a serverless and containerized computing service - Through Virtual Kubelet technology and the Volcano Engine container service VKE, the elastic container instance VCI can provide users with flexible computing and Kubernetes orchestration functions, supporting Fast startup, high-concurrency creation, and secure sandbox container isolation.

By integrating the advantages of Serverless and containers, Elastic Container Instance (VCI) can provide enterprises with an efficient, reliable and secure operating environment, allowing users to focus on developing and running cloud-native architecture applications without having to worry about The management and maintenance of the underlying infrastructure, and only paying for the resources required for actual business operation, help enterprises more reasonably control cloud costs.

No operation and maintenance of infrastructure, no planning of computing resources

In the traditional self-built Kubernetes cluster architecture, whether it is the control plane or the data plane, as a container group Pod The running carrier and nodes are located at the core of the Kubernetes architecture. When node computing resources are insufficient, computing resources need to be supplemented by adding nodes to support the deployment or expansion of business workloads.

On the cloud, various managed Kubernetes services represented by Volcano Engine Container Service VKE will replace users in managing and operating the Kubernetes control plane, providing managed services with SLA guarantees. For the data plane that runs users' actual business workloads, users usually need to operate and maintain working nodes by themselves, and the computing resources provided by the nodes need to be planned in advance according to business characteristics. For example, when business peaks require advance expansion, Business downturns require downsizing.

Therefore, in the traditional node-centered Kubernetes architecture, node operation and maintenance often requires a lot of energy from the technical team; at the same time, how to balance the issue of advanced planning of resources and resource costs has also become a "refined application for enterprises. Questions that we have to think about after the era of "cloud ".

Elastic Container Instance VCI provides Serverless and containerized computing services. Each VCI will only provide the operating environment and computing resources for a single Pod. Users do not need to care about the operation and maintenance and capacity planning of the underlying nodes. They only need to Just deploy the container application. When using VCI to deploy container applications, you can use native Kubernetes semantics, which is no different from the node center architecture.

At the resource level, the elastic container instance VCI uses the Volcano Engine computing infrastructure as a resource pool to provide container computing resources of multiple resource types to facilitate users to flexibly choose according to business needs. At the same time, users do not need to pay attention to resource capacity planning issues. VCI provides massive computing resources and will not cause phenomena such as Pod drift and rescheduling that are common in traditional node-centric architectures, which will further cause service interruptions, performance fluctuations, Problems such as data inconsistency and scheduling delays.

Finally, the seamless integration of container service VKE and elastic container instance VCI not only realizes full hosting of the Kubernetes cluster control plane, but also provides users with full Serverless data plane (i.e. elastic container instance VCI). Hosting, operation-free, and secure isolation greatly reduce users’ operation and maintenance costs on Kubernetes infrastructure and eliminate the complexity of capacity planning, allowing users to focus more on their business applications.

Ultimate elasticity, you get what you need

Flexibility and savings

According to Flexera in the "2022 State of the Cloud Report" According to the disclosed survey data, the self-assessed cloud cost waste of the interviewed companies reached 32%, and "cost" has become one of the cloud challenges that companies are most concerned about for three consecutive years - "refined cloud usage" has begun to gain traction. More and more enterprises are paying attention to how to save cloud costs while meeting business development needs, which has become an important challenge that enterprises need to solve when using the cloud.

Elastic Container Instance VCI focuses on the enterprise's "cloud cost control" needs from the beginning of its product design, emphasizing refined billing based on the actual amount of resources used, supporting multiple billing models, and helping enterprises truly make good use of the cloud. .

Refined billing: Elastic container instance VCI is billed according to the resources actually used by the user container group (Pod). Billing resources include vCPU, memory, GPU, etc. At the same time, the billing time of the VCI instance is its running time, which is the time used by the user Pod from downloading the container image to stopping running. It is accurate to the second level, truly realizing billing based on actual usage.

Improve the packing rate: Under the cloud native architecture, compared with traditional computing resources, another advantage of elastic container instances is to reduce idle resources and improve the packing rate, thereby reducing user costs. computing resource usage costs. Specifically, when a Kubernetes cluster uses cloud server ECS as a working node, it needs to run the system components required by Kubernetes on the node. Therefore, some computing resources need to be reserved, and there is a certain additional cost; and if an elastic container instance VCI is used as the computing resource , you only need to pay for the resources actually used by the business Pod:

  • Example: Running 3 business Pods, their resource requirements for vCPU are divided into 16C, 8C, 8C
  • Use ECS: At least 2 32C vCPU ECS nodes are required (resources need to be reserved for the system on each node)
  • Use VCI: Create 3 VCI instances equal to the resources required by the Pod, that is, 32 C computing resources (no resources are idle)

火山引擎弹性容器实例:从节点中心转型 Serverless 化架构的利器

Rich billing methods: The current elastic container instance VCI has officially supported pay-as-you-go billing, and will launch an invitation test event for the preemptive instance (Spot) payment model, as well as reserved instances and flexible reservations. Prepaid models such as instances are also supported by plans. By providing multiple payment models, we hope to help users further optimize cost management and budgeting of computing resources, and choose different payment models for different business scenarios, thereby better coordinating business resource needs and cloud cost planning.

Bounce out

Elastic Container Instance VCI fully integrates the computing resource infrastructure of Volcano Engine to provide computing power, and provides a variety of instance specification families for users to choose based on different underlying hardware capabilities, providing differentiated computing, storage, and network performance. , supporting coverage of a variety of business applications and service scenarios.

CPU General-purpose specification family: Such as general-purpose computing u1, general-purpose n3i, etc., which provide balanced vCPU, memory, and network capabilities to meet the needs of large-scale users. Service requirements in most scenarios.

GPU Computing type specification family: Such as GPU computing type gni2 (equipped with NVIDIA A10 GPU graphics card), GPU computing type g1v (equipped with NVIDIA V100 GPU graphics card), etc. It can provide a cost-effective use experience in various AI computing scenarios such as large model training, text and image generation, and task reasoning.

The latest general computing type u1 Instance specification family relies on the Volcano engine resource pooling technology and intelligent scheduling algorithm for dynamic resource management , can provide enterprises with a stable supply of computing power, and support a variety of processors and flexible processor memory ratios, from fine-grained small specifications such as 0.25C-0.5Gi, 0.5C-1Gi, 1C-2Gi to 24C-48Gi , 32C-256Gi and other large specifications are available for selection, with extremely high cost performance!

Backed by a huge supply of computing resources, the elastic container instance VCI also provides users with an industry-leading elastic resource priority scheduling strategy : Combined with the container service VKE, we support users to use their own Define the resource policy (ResourcePolicy) to elastically schedule to different types of computing resource pools such as cloud servers ECS and elastic container instances VCI, and perform scheduling and allocation according to percentage thresholds, priority control and other strategies according to actual business scenarios.

This means that users can combine the cloud server ECS computing resource pool and the elastic container instance VCI computing resource pool to enjoy massive Volcano Engine computing resources; at the same time, combine the Cluster Autoscaler capabilities and VCI in the container cluster , even if computing resources are not requested in advance in the container cluster, users can obtain elastic resources in a timely manner to efficiently support business applications.

bounce fast

Faced with unpredictable business traffic peaks in online business, it is crucial to ensure stable business operation and user experience; at the critical moment of business development, Quickly popping up computing resources and quickly enabling business applications are also critical. Elastic container instances are also naturally suitable for these scenarios.

火山引擎弹性容器实例:从节点中心转型 Serverless 化架构的利器

On the one hand, the elastic capability of the Volcano Engine Elastic Container Instance (VCI) can meet the needs of tens of thousands of cores per minute vCPU computing resources, ensuring that When business needs it, sufficient computing resources are quickly available. When the traffic peak is over and the business workload decreases, elastic computing resources can be quickly released, making cloud costs more economical.

On the other hand, container image pulling often takes a long time, which is even more serious when large-scale Pods are started concurrently. Elastic container instance VCI not only has a variety of container image acceleration capabilities provided by VKE, such as container image lazy loading (image lazy loading), P2P container image distribution, etc., but also has unique acceleration capabilities such as container image caching capabilities. According to actual test data, Container image caching can control the overall startup time of a Pod with a container image (data volume within 100 G) to the level of ten seconds, helping users greatly improve efficiency and reduce cloud costs.

Conclusion

Since its official launch in 2022, Elastic Container Instance VCI has served many large and medium-sized enterprise customers and has been fully utilized in Bytedance’s internal and external business scenarios. practical test. The Volcano Engine cloud native team also continuously polishes product capabilities based on customer service experience and continues to improve the quality and stability of product services:

  • In terms of ecological integration, Elastic Container Instance (VCI) not only continues to enhance its integration with the Kubernetes ecosystem, but can also be seamlessly integrated with the Volcano Engine Container Service (VKE), making it easier for customers to deploy and manage containers in a cloud-native environment. Application;
  • In terms of distributed container image startup acceleration, VCI collaborates with Volcano Engine Image Warehouse CR and other products to improve the Dragonfly architecture, further enhancing the efficiency and reliability of image distribution to VCI;
  • VCI's unique container image cache acceleration capability better solves the problems of slow Pod startup and low image pull success rate in large-scale concurrency and large container image scenarios, and reduces the pressure on the image warehouse;
  • In terms of diversity of computing resource specifications, VCI will launch more optional resources in 2023, such as n3i, a new general-purpose computing instance based on Volcano Engine’s self-developed DPU fully offloaded architecture, which shields back-end server differences and has unlimited theoretical capabilities The general computing power instance u1 of the resource pool, etc.

The Volcano Engine cloud native team is mainly responsible for the construction of the PaaS product system in the Volcano Engine public cloud and privatized scenarios. Combined with ByteDance’s many years of cloud native technology stack experience and best practices, it helps Enterprises accelerate digital transformation and innovation. Products include container services, image warehouses, distributed cloud native platforms, function services, service grids, continuous delivery, observable services, etc.

Facing 2024, Elastic Container Instance VCI will continue to improve and explore under the Serverless Kubernetes product architecture and form, providing more internal and external customers with no operation and maintenance of infrastructure, no planning of computing resources, and ultimate flexibility. Serverless containerized products and services can better support customers' various GPU and CPU computing power needs in AIGC, bioinformatics scientific computing, social e-commerce and other business scenarios.

The above is the detailed content of Volcano Engine elastic container instance: a powerful tool for transforming from a node-centric architecture to a serverless architecture. For more information, please follow other related articles on the PHP Chinese website!

Statement of this Website
The content of this article is voluntarily contributed by netizens, and the copyright belongs to the original author. This site does not assume corresponding legal responsibility. If you find any content suspected of plagiarism or infringement, please contact admin@php.cn

Hot AI Tools

Undresser.AI Undress

Undresser.AI Undress

AI-powered app for creating realistic nude photos

AI Clothes Remover

AI Clothes Remover

Online AI tool for removing clothes from photos.

Undress AI Tool

Undress AI Tool

Undress images for free

Clothoff.io

Clothoff.io

AI clothes remover

Video Face Swap

Video Face Swap

Swap faces in any video effortlessly with our completely free AI face swap tool!

Hot Article

Roblox: Bubble Gum Simulator Infinity - How To Get And Use Royal Keys
3 weeks ago By 尊渡假赌尊渡假赌尊渡假赌
Nordhold: Fusion System, Explained
3 weeks ago By 尊渡假赌尊渡假赌尊渡假赌
Mandragora: Whispers Of The Witch Tree - How To Unlock The Grappling Hook
3 weeks ago By 尊渡假赌尊渡假赌尊渡假赌

Hot Tools

Notepad++7.3.1

Notepad++7.3.1

Easy-to-use and free code editor

SublimeText3 Chinese version

SublimeText3 Chinese version

Chinese version, very easy to use

Zend Studio 13.0.1

Zend Studio 13.0.1

Powerful PHP integrated development environment

Dreamweaver CS6

Dreamweaver CS6

Visual web development tools

SublimeText3 Mac version

SublimeText3 Mac version

God-level code editing software (SublimeText3)

Hot Topics

Java Tutorial
1665
14
PHP Tutorial
1269
29
C# Tutorial
1249
24
DeepMind robot plays table tennis, and its forehand and backhand slip into the air, completely defeating human beginners DeepMind robot plays table tennis, and its forehand and backhand slip into the air, completely defeating human beginners Aug 09, 2024 pm 04:01 PM

But maybe he can’t defeat the old man in the park? The Paris Olympic Games are in full swing, and table tennis has attracted much attention. At the same time, robots have also made new breakthroughs in playing table tennis. Just now, DeepMind proposed the first learning robot agent that can reach the level of human amateur players in competitive table tennis. Paper address: https://arxiv.org/pdf/2408.03906 How good is the DeepMind robot at playing table tennis? Probably on par with human amateur players: both forehand and backhand: the opponent uses a variety of playing styles, and the robot can also withstand: receiving serves with different spins: However, the intensity of the game does not seem to be as intense as the old man in the park. For robots, table tennis

The first mechanical claw! Yuanluobao appeared at the 2024 World Robot Conference and released the first chess robot that can enter the home The first mechanical claw! Yuanluobao appeared at the 2024 World Robot Conference and released the first chess robot that can enter the home Aug 21, 2024 pm 07:33 PM

On August 21, the 2024 World Robot Conference was grandly held in Beijing. SenseTime's home robot brand "Yuanluobot SenseRobot" has unveiled its entire family of products, and recently released the Yuanluobot AI chess-playing robot - Chess Professional Edition (hereinafter referred to as "Yuanluobot SenseRobot"), becoming the world's first A chess robot for the home. As the third chess-playing robot product of Yuanluobo, the new Guoxiang robot has undergone a large number of special technical upgrades and innovations in AI and engineering machinery. For the first time, it has realized the ability to pick up three-dimensional chess pieces through mechanical claws on a home robot, and perform human-machine Functions such as chess playing, everyone playing chess, notation review, etc.

Claude has become lazy too! Netizen: Learn to give yourself a holiday Claude has become lazy too! Netizen: Learn to give yourself a holiday Sep 02, 2024 pm 01:56 PM

The start of school is about to begin, and it’s not just the students who are about to start the new semester who should take care of themselves, but also the large AI models. Some time ago, Reddit was filled with netizens complaining that Claude was getting lazy. "Its level has dropped a lot, it often pauses, and even the output becomes very short. In the first week of release, it could translate a full 4-page document at once, but now it can't even output half a page!" https:// www.reddit.com/r/ClaudeAI/comments/1by8rw8/something_just_feels_wrong_with_claude_in_the/ in a post titled "Totally disappointed with Claude", full of

At the World Robot Conference, this domestic robot carrying 'the hope of future elderly care' was surrounded At the World Robot Conference, this domestic robot carrying 'the hope of future elderly care' was surrounded Aug 22, 2024 pm 10:35 PM

At the World Robot Conference being held in Beijing, the display of humanoid robots has become the absolute focus of the scene. At the Stardust Intelligent booth, the AI ​​robot assistant S1 performed three major performances of dulcimer, martial arts, and calligraphy in one exhibition area, capable of both literary and martial arts. , attracted a large number of professional audiences and media. The elegant playing on the elastic strings allows the S1 to demonstrate fine operation and absolute control with speed, strength and precision. CCTV News conducted a special report on the imitation learning and intelligent control behind "Calligraphy". Company founder Lai Jie explained that behind the silky movements, the hardware side pursues the best force control and the most human-like body indicators (speed, load) etc.), but on the AI ​​side, the real movement data of people is collected, allowing the robot to become stronger when it encounters a strong situation and learn to evolve quickly. And agile

ACL 2024 Awards Announced: One of the Best Papers on Oracle Deciphering by HuaTech, GloVe Time Test Award ACL 2024 Awards Announced: One of the Best Papers on Oracle Deciphering by HuaTech, GloVe Time Test Award Aug 15, 2024 pm 04:37 PM

At this ACL conference, contributors have gained a lot. The six-day ACL2024 is being held in Bangkok, Thailand. ACL is the top international conference in the field of computational linguistics and natural language processing. It is organized by the International Association for Computational Linguistics and is held annually. ACL has always ranked first in academic influence in the field of NLP, and it is also a CCF-A recommended conference. This year's ACL conference is the 62nd and has received more than 400 cutting-edge works in the field of NLP. Yesterday afternoon, the conference announced the best paper and other awards. This time, there are 7 Best Paper Awards (two unpublished), 1 Best Theme Paper Award, and 35 Outstanding Paper Awards. The conference also awarded 3 Resource Paper Awards (ResourceAward) and Social Impact Award (

Li Feifei's team proposed ReKep to give robots spatial intelligence and integrate GPT-4o Li Feifei's team proposed ReKep to give robots spatial intelligence and integrate GPT-4o Sep 03, 2024 pm 05:18 PM

Deep integration of vision and robot learning. When two robot hands work together smoothly to fold clothes, pour tea, and pack shoes, coupled with the 1X humanoid robot NEO that has been making headlines recently, you may have a feeling: we seem to be entering the age of robots. In fact, these silky movements are the product of advanced robotic technology + exquisite frame design + multi-modal large models. We know that useful robots often require complex and exquisite interactions with the environment, and the environment can be represented as constraints in the spatial and temporal domains. For example, if you want a robot to pour tea, the robot first needs to grasp the handle of the teapot and keep it upright without spilling the tea, then move it smoothly until the mouth of the pot is aligned with the mouth of the cup, and then tilt the teapot at a certain angle. . this

Distributed Artificial Intelligence Conference DAI 2024 Call for Papers: Agent Day, Richard Sutton, the father of reinforcement learning, will attend! Yan Shuicheng, Sergey Levine and DeepMind scientists will give keynote speeches Distributed Artificial Intelligence Conference DAI 2024 Call for Papers: Agent Day, Richard Sutton, the father of reinforcement learning, will attend! Yan Shuicheng, Sergey Levine and DeepMind scientists will give keynote speeches Aug 22, 2024 pm 08:02 PM

Conference Introduction With the rapid development of science and technology, artificial intelligence has become an important force in promoting social progress. In this era, we are fortunate to witness and participate in the innovation and application of Distributed Artificial Intelligence (DAI). Distributed artificial intelligence is an important branch of the field of artificial intelligence, which has attracted more and more attention in recent years. Agents based on large language models (LLM) have suddenly emerged. By combining the powerful language understanding and generation capabilities of large models, they have shown great potential in natural language interaction, knowledge reasoning, task planning, etc. AIAgent is taking over the big language model and has become a hot topic in the current AI circle. Au

Hongmeng Smart Travel S9 and full-scenario new product launch conference, a number of blockbuster new products were released together Hongmeng Smart Travel S9 and full-scenario new product launch conference, a number of blockbuster new products were released together Aug 08, 2024 am 07:02 AM

This afternoon, Hongmeng Zhixing officially welcomed new brands and new cars. On August 6, Huawei held the Hongmeng Smart Xingxing S9 and Huawei full-scenario new product launch conference, bringing the panoramic smart flagship sedan Xiangjie S9, the new M7Pro and Huawei novaFlip, MatePad Pro 12.2 inches, the new MatePad Air, Huawei Bisheng With many new all-scenario smart products including the laser printer X1 series, FreeBuds6i, WATCHFIT3 and smart screen S5Pro, from smart travel, smart office to smart wear, Huawei continues to build a full-scenario smart ecosystem to bring consumers a smart experience of the Internet of Everything. Hongmeng Zhixing: In-depth empowerment to promote the upgrading of the smart car industry Huawei joins hands with Chinese automotive industry partners to provide

See all articles