Content

Target customers: Internet enterprises that have high requirements to deploy workloads in large-scale production environments. The requirements include high security, stability, and observability. Big data computing enterprises that run clusters with high stability, performance, and efficiency to implement large-scale data computing, high-performance data processing, and high elasticity. International enterprises that attach great importance to security, privacy, and SLAs with indemnification clauses. This applies when enterprises run a business in China. Financial enterprises and governments that require high security protections and SLAs with indemnification clauses. Features released: reliable managed master nodes: Each API server supports auto scaling and ensures smooth scaling of clusters with a large number of nodes. etcd supports disaster recovery, backup, and manual recovery. etcd uses the hot standby mechanism to ensure the availability of cluster databases. Key metrics of control components are visualized to analyze results and prevent risks. Secure clusters: In the control plane, etcd uses encrypted storage by default. In the data plane, the KMS plug-in can be installed to enable secrets encryption at rest. Security management is supported to enable comprehensive security detection and automatic recovery for running containers. Intelligent scheduling: kube-scheduler provides enhanced scheduling performance and supports multiple intelligent scheduling algorithms. It can be used to schedule neural processing unit (NPU) resources. It can also be used to optimize container scheduling capabilities in scenarios, such as large-scale data computing and high-performance data processing. SLA guarantee: An SLA that includes indemnification clauses is supported to guarantee 99.95% availability for each API server.