Abstract: CNCF (Cloud Native Computing Foundation) officially accepted the multi-cloud container choreography project Karmada contributed by Huawei Cloud, ushering in the first multi-cloud container choreography project of CNCF.

On September 15, Beijing time, CNCF (Cloud Native Computing Foundation) officially accepted the multi-cloud container Choreography project contributed by Huawei Cloud Karmada (github.com/karmada-io/…). , ushered in CNCF’s first multi-cloud container choreography project. The addition of Karmada project further expands CNCF’s cloud native map to the distributed cloud field.

Huawei Cloud has been actively giving back to the community in terms of technology. It has opened a series of cloud native projects represented by intelligent edge project KubeEdge and batch computing project Volcano. Karmada project is jointly initiated by 8 enterprises including Huawei Cloud, INDUSTRIAL and Commercial Bank of China, Xiaohongshu and FaW, which precipates the rich accumulation of various enterprises in the field of cloud management and provides detailed and effective practical guidance and help for developers. Using Karmada, an infinite and scalable container resource pool can be built. Let developers use multi-cloud clusters like a single Kubernetes cluster.

Karmada introduction

With the rapid development of enterprise services, multi-cloud has gradually become the infrastructure for data center construction. Scenarios such as multi-region DISASTER recovery and multi-activity, large-scale multi-cluster management, cross-cloud flexibility and migration have promoted the rapid development of cloud native multi-cloud related technologies. However, in the actual production process, the cloud native cloud still faces the following challenges:

  • Repeated work in multiple clusters: O&M engineers need to deal with complex cluster configuration, management differences between clusters of different cloud vendors, and fragmented API access points.

  • The maintenance problem of scattered services is as follows: the differential configuration of applications in different clusters is cumbersome; Service access across clouds and application synchronization between clusters are difficult to manage.

  • Cluster boundaries: Application availability is limited by the cluster; Resource scheduling and elastic scaling are limited to clusters.

  • Vendor binding: Stickiness of service deployment, lack of automatic failover; Lack of a neutral open source cloud container choreography project.

Karmada combines the core practices of Huawei Cloud and Multi-cloud container platform MCP and Kubernetes Federation, and integrates many new technologies: Including Kubernetes native API support, multi-tier high availability deployment, multi-cluster automatic failover, multi-cluster application automatic scaling, multi-cluster service discovery, etc., and provide native Kubernetes smooth evolution path, so that Karmada based multi-cloud solution seamlessly into the cloud native technology ecology, Provides enterprises with a smooth evolution scheme from a single cluster to a multi-cloud architecture.

Overview of Karmada project

Ecological cooperation

Karmada project was jointly initiated by Huawei Cloud, INDUSTRIAL and Commercial Bank of China, Shanghai Pudong Development Bank, Xiaohongshu, VIPKID, Qutoutiao, China FaW and T3 Travel, and officially announced open source at Huawei Developer Conference (HDC.Cloud) 2021 on April 25, 2021. Karmada has received wide attention and support since open source. At present, more than 30 large enterprises/institutions/universities have participated in community development and contribution.

Karmada MCP project from huawei cloud cloudy container platform, at the same time into the industrial and commercial bank, the little red book, China faw and other clients of different industries in cloudy management experience, can provide companies with detailed and effective landing guidance and help enterprises through Karmada building across different cloud, data center promise extensible application resource pool, It can manage clusters and applications in different clouds and data centers as easily and conveniently as managing a single Kubernetes cluster. — Zhang Yuxin, CTO of Huawei Cloud

In terms of community contribution, ICBC, as the head participant of Karmada project, combined with icbc’s experience in multi-container cluster management for many years, has carried out in-depth customized development of core modules such as cluster lifecycle management and core scheduling controller. In the future, ICBC will continue to participate in the development and management of Karmada community, and plan to continue in-depth research and contribution in modules such as multi-cluster automatic scheduling and multi-cluster automatic scaling, so as to feed back the open source community and continue to expand its influence in the industry. — Lu Jinbiao, an expert at ICBC software Development Center

Complex scenarios of multiple mixed containers exist in group enterprises at the same time, and the operation and maintenance problems across the multi-cloud technology architecture are increasingly prominent. As the centralized management architecture of multi-cluster, multi-cloud and hybrid cloud, compatible with native Kubernetes API, Huawei Karmada effectively solves the current pain points such as container arrangement of multi-cluster, multi-environment can not be centralized management, security isolation mechanism is not sound. It is hoped that Karmada will continue to expand its functions through the joint maintenance and contribution of the community after joining CNCF. I look forward to Karmada’s graduation from CNCF as soon as possible and return to the original ecology of cloud. — Wang Guang, Technical Operation Director of Digital Department of Automobile System

Karmada’s native compatibility with the Kubernetes API allows for unmodified docking with the existing Kubernetes ecosystem. In the implementation process, we used Karmada to connect with the existing GitOps ecosystem, which greatly improved the efficiency of application cross-cluster deployment. — VIPKID Operation and Maintenance Director Gu Yuhu

Karmada provides rich multi-cluster scheduling policies and built-in policy sets out of the box, which can greatly simplify the system complexity in the two-site, three-center, remote DISASTER recovery, and same-city active-active architectures, which are critical to the financial industry. — Lu Binggang, Director of Cloud Transformation Department of SPD Bank

It should be a future

At present, Karmada has been commercially used in Huawei Multi-Cloud Container Platform (MCP) to provide distributed Cloud solutions and key capabilities such as unified multi-cluster management, unified application deployment, and traffic distribution across clouds.

In addition to MCP, Karmada has been established in dozens of enterprises from finance, Internet, education and so on.

CNCF officially accepted Karmada as the first multi-cloud container arrangement project in the field of cloud native, which will greatly promote the ecological construction and cooperation of Karmada upstream and downstream communities and attract the deep participation of cloud native enterprise users. Karmada will play an increasingly important role in the fields of multi-cluster application management, service governance, high availability deployment and so on. Huawei Cloud will continue to work in the cloud native field, continue to lead innovation, prosperity of the ecosystem, and help all industries to the road of rapid and intelligent development.

Click to follow, the first time to learn about Huawei cloud fresh technology ~