Huawei Cloud CCE Autopilot Pioneers Automated Era

Huawei Cloud CCE Autopilot Pioneers Automated Era

On June 22, 2024, during the HDC.2024 session All Serverless: Innovative Technology Empowers Industry Practices, Ruan Binfeng gave a speech titled The Cost-Effective and Agile Path of Serverless Containers. He emphasized that serverless computing has become the most popular form of cloud native computing. Huawei Cloud CCE Autopilot and CCI enable serverless workloads to run in scenarios like AI, HPC, and web to improve user experience.

So let's talk about CCE Autopilot. Its O&M-free functionality optimizes the Kubernetes compatibility of container services, freeing up time to focus on developing and implementing core services.



Huawei Cloud Is Committed to Serverless Infrastructure, Simplifying O&M Management, and Unleashing Innovation Potential

In cloud computing, container technologies have become a strong driver for enterprise IT architecture transformation, with their lightweight nature and high efficiency. However, traditional container services that rely on the serverful infrastructure are revealing the following issues, which severely hinder the pace of enterprise innovation:

l?? O&M management: Enterprises need to manually manage server resource allocation and expansion, which involves complex capacity planning and resource scheduling, as well as continuous O&M such as node monitoring, troubleshooting, and system upgrades. This is expensive and requires a large workforce and many resources.

l?? Scalability: Enterprises need to create joint scaling policies for nodes and workloads, but worker nodes need to be scaled in advance, which takes a few minutes and affects efficiency and response speed.

l?? Cost control: Enterprises need to allocate resources to nodes in advance. Unfortunately, those resources are often underutilized, or when there are too many workloads or the workloads are too large, resources may be insufficient. This makes it hard to maximize cost effectiveness.

Huawei Cloud has a deep understanding of the pain points that industries are experiencing and is dedicated to addressing these pain points by developing and innovating serverless infrastructure to provide comprehensive and optimized workload support. To address the O&M challenges of independent resource pools, Huawei Cloud developed CCE Autopilot. CCE Autopilot uses a converged resource base for serverless compute. This allows for the pooling of resources such as CPUs, memory, and GPUs, which improves resource flexibility and scalability. With this converged resource base, you do not need to manage node workloads and resource pool carrying capacity. With CCE Autopilot, a converged resource pool centrally manages resources. You just need to manage the container lifecycle. This provides a flexible and easy-to-manage computing environment, which simplifies O&M management and unleashes innovation potential.



CCE Autopilot: Kubernetes Cluster Hosting Accelerates Application Modernization

CCE Autopilot is a serverless container service that provides enhanced compatibility with Kubernetes and frees you from tiring O&M tasks. Traditionally, Huawei Cloud only handles master nodes. You would still have to manage all the worker nodes and their applications. To effectively maintain and scale containerized applications, you needed to be very familiar with Kubernetes. O&M of the master nodes has since been automated. Huawei Cloud handles both the master and worker nodes. You can deploy your applications without purchasing nodes. You just worry about implementing the application service logic. This new mode of operations greatly reduces O&M costs and improves application reliability and scalability.

Compared with CCE or CCE Turbo clusters, CCE Autopilot clusters have the following advantages:

l?? Serverless evolution: Worker nodes are fully hosted on Huawei Cloud, so you do not have to maintain node deployment, manage nodes, or worry about security issues. Cluster specifications are automatically scaled.

l?? Resource pooling: A serverless converged resource pool is used to manage resources such as CPUs, memory, and GPUs, reducing resource fragments and enabling on-demand use of container resources.

l?? Performance optimization: The resource pool is pre-warmed to provision resources faster, so more containers can be scaled within seconds based on the size of a given workload.



Key Capabilities of CCE Autopilot

CCE Autopilot redefines the management and O&M of container services with key capabilities, making it much easier for you to handle service challenges and speed up innovation.



Intelligent, Reliable, and Automated O&M

CCE Autopilot provides you with more stable, secure, and intelligent cluster experience through intelligent version upgrades, automatic vulnerability fixing, and intelligent parameter tuning. As a serverless solution, CCE Autopilot simplifies capacity planning and node purchase processes of fully hosted clusters. You do not need to manage or maintain the underlying resources, so there is much less O&M to deal with. You can just focus on developing and deploying service logic.

Excellent Scalability with Continuous Iterations

CCE Autopilot provides excellent scalability by providing a serverless resource base. Multi-level resource pool warm-up can meet heterogeneous resource requirements. CCE Autopilot also improves resource scalability through continuous iterations of the resource pool, so you do not need to plan or reserve resources in advance to handle traffic bursts, seasonal fluctuations, and long-term growth. Containers can be scaled in or out within seconds to adapt to changing workloads, which ensures service continuity and excellent performance. With CCE Autopilot, you can quickly launch new applications or services in response to market changes.

Compatibility with Cloud Native Open Source Ecosystem

CCE Autopilot leverages a serverless architecture and a cloud native open source ecosystem to provide serverless clusters compatible with the Kubernetes ecosystem. This allows you to flexibly expand functions as needed. CCE Autopilot can also keep up with all Kubernetes community versions, so you always get the latest technical updates and security patches in a timely manner. It makes it easier to keep current with cutting-edge technologies. CCE Autopilot is always integrating mainstream open source security, application management, scalability, CI/CD, and AI software, such as OPA Gatekeeper and Knative. Integrating all these tools gets you an out-of-the-box application framework for easier application management.

Flexible Specifications and Per-Second Billing

CCE Autopilot delivers a flexible, efficient, and cost-effective cloud service experience. It automatically and dynamically adjusts cluster specifications and eliminates traditional specification tier restrictions. With CCE Autopilot, you can flexibly adjust specifications starting from as little as 0.25 vCPUs, and you can configure whatever resource ratio is required. The pay-per-use billing of CCE Autopilot (measured in seconds) means you save money because you only pay for the resources you actually use.

Security Isolation and Automatic Warning

CCE Autopilot, built on the QingTian architecture, improves application security through virtual machine isolation. It provides a simplified and secure running environment using a dedicated container OS. With the underlying resource pool, CCE Autopilot supports quick fault isolation and rectification, ensuring continuous, secure, and stable application performance. Built-in automatic warnings can identify and prevent overloads on the control plane in a timely manner. The control plane components can be automatically scaled out to handle extra loads, ensuring service stability and reliability.

Application Scenarios

CCE Autopilot provides serverless infrastructure, which makes it suitable for a variety of service scenarios.



O&M and Iterations of SaaS/Enterprise Platforms

CCE Autopilot is the best choice for SaaS and enterprise platforms, especially for enterprises who have large resource pools that need to be iterated frequently. Traditionally, you need to handle your own O&M and upgrades, which drives labor costs through the roof. The automated O&M of CCE Autopilot puts an end to this issue. Internet finance enterprises have demanding compliance requirements. Traditionally, it is hard to develop compliance capabilities for OSs. CCE Autopilot simplifies node management and improves system security and compliance. This enables you to focus better on the innovation and development of core services.

Efficient, Auto Scaling for Services

For Internet entertainment, social networking, and ride-hailing, CCE Autopilot provides auto scaling to dynamically adjust resource configurations based on service characteristics and predicted traffic during traffic bursts. Unlike traditional scheduled scaling, CCE Autopilot scaling ensures efficient matching of resources to service requirements. This type of auto scaling allows you to optimize your enterprise's cost structure and reduce resource waste while still maintaining service continuity.

Cost Optimization Configuration

CCE Autopilot provides flexible resource configuration solutions that simplify learning processes and help reduce costs. It also supports auto scaling of services to help enterprises keep pace with business growth. With CCE Autopilot, start-ups can enjoy high-performance and reliable services, even when their resource needs are minimal. As their services grow, resources can be scaled to ensure cost-effectiveness and service continuity.


How Do I Use CCE Autopilot Clusters?

https://www.huaweicloud.com/intl/en-us/product/cce.html

Step 1???? Log in to the Huawei Cloud management console, click Service List in the upper left corner, and choose Cloud Container Engine.

Step 2??? On the CCE console, choose Buy Cluster, select CCE Autopilot cluster, and configure basic network add-ons to create a CCE Autopilot cluster.

?Step 3??? Go to the cluster console, where you can access functions such as Kubernetes resource management. Unlike CCE clusters, there is no need to manage nodes and node pools of CCE Autopilot clusters. Huawei Cloud handles all the infrastructure hosting. Nonetheless, you can still view all of the Kubernetes control plane details.


Huawei Cloud CCE Autopilot takes container services to the next level with its serverless architecture. It provides O&M-free, dynamic resource management, intelligent prediction, and automated O&M, which makes IT management simpler, reduces operation costs, and improves resource utilization. CCE Autopilot is also compatible with the Kubernetes ecosystem, which helps support seamless interconnection and agile market adaptation. We invite you to join hands with us to usher in a new chapter of enterprise digital transformation and embrace a more efficient, more intelligent future.

要查看或添加评论,请登录

社区洞察

其他会员也浏览了