What's New

1

CCE Autopilot clusters are in commercial use.

CCE Autopilot allows you to create serverless clusters that offer optimized Kubernetes compatibility and free you from O&M. After a CCE Autopilot cluster is created, you can deploy applications without purchasing nodes or maintaining the deployment, management, and security of nodes. You only need to focus on the implementation of application service logic, which greatly reduces your O&M costs and improves the reliability and scalability of applications.

Commercial use

What Is a CCE Autopilot Cluster?

2

CCE supports FlexusX.

You are allowed to choose FlexusX when creating nodes in CCE clusters.

Commercial use

Node Specifications

3

Expanding the capacity or modifying the specifications of the disk attached to a CCE node does not impact the process of resetting the node.

When you increase the capacity or modify the specifications of a disk attached to a CCE node through the ECS console, the node can be reset on the CCE console.

Commercial use

Resetting a Node

4

CCE enables you to designate directories when you are creating SFS 3.0 file systems using the StorageClass objects.

When creating SFS 3.0 file systems using the StorageClass objects, you can specify directories.

Commercial use

Creating an SFS Subdirectory Using a Dynamic PV

5

Custom node pools in CCE standard or Turbo clusters can accept nodes from the default node pool.

If the node pool constraints are met, you can migrate nodes from the default node pool to an on-premises node pool to fulfill various service needs and enhance resource utilization.

Commercial use

Migrating a Node

6

CCE allows you to customize the default image path of containerd.

You can customize the default image path of containerd on a node and choose image sources as required.

Commercial use

Modifying Node Pool Configurations

1

CCE offers specific tags for unsubscribed nodes.

By adding a specific tag to a node when unsubscribing, alarm rules can be configured for an on-premises Prometheus platform to prevent false alarms.

Commercial use

Deleting or Unsubscribing from a Node

2

Kubernetes 1.29 is now commercially available in CCE.

CCE allows you to create clusters of Kubernetes 1.29.

Commercial use

Kubernetes 1.29 Release Notes

3

CCE enables you to cancel node drainage.

You can cancel the node drainage. Once the drainage is canceled, scheduling is restored on the nodes.

Commercial use

Draining a Node

4

CCE allows you to retain labels and taints when a node is reset.

You have the option to keep or delete specific metadata, including labels and taints, when resetting a node in the default node pool.

Commercial use

Resetting a Node

5

CCE allows you to update node pool agencies.

When updating a node pool, you have the option to adjust the user agency settings associated with the node pool. This allows for more flexible management of access and operation permissions for the node pool.

Commercial use

Updating a Node Pool

6

CCE supports custom header forwarding policies for ingresses.

Dedicated load balancer ingresses support custom header forwarding policies. You can configure different header key-value pairs to determine the backend Service to which data is forwarded.

Commercial use

Configuring a Custom Header Forwarding Policy for a LoadBalancer Ingress

7

CCE has upgraded node pools, which now offer additional benefits of pay-per-use and yearly/monthly billing modes.

Upgraded node pools provide additional benefits beyond flexible payment options (pay-per-use and yearly/monthly). They also enhance configuration management and improve resource management efficiency and flexibility.

Commercial use

Upgrading to New Node Pools

8

CCE has optimized the node drainage failure message.

In the event that a node cannot be drained, a clear and detailed error message will be shown.

Commercial use

Draining a Node

9

CCE has increased the maximum number of Elastic Volume Service (EVS) disks that can be attached to each node.

The ability to attach disks has been significantly improved, allowing for up to 16 EVS disks to be attached to a single node.

Commercial use

Creating a Node

10

CCE has improved the allocation of space for container engines, allowing for the mounting of space to specific paths.

When creating a node, if the total allocated space for the container engine, container images, kubelet, and emptyDir is less than 100%, you can mount the remaining space to a specified path for custom data storage.

Commercial use

Creating a Node

11

CCE supports the use of existing EIPs for pods in CCE Turbo clusters.

When setting up an EIP for a pod in a CCE Turbo cluster, you have the option to use an existing EIP.

Commercial use

Configuring an EIP for a Pod

12

CCE allows for the replication of default alarm rules.

The default CCE alarm rules can be replicated and modified to fit your needs.

Commercial use

Configuring Custom Alarms on CCE

1

CCE supports yearly/monthly or pay-per-use billing mode for EVS disks that are dynamically created.

When dynamically creating EVS disks, you can select yearly/monthly or pay-per-use billing mode for the disks.

Commercial use

Using an EVS Disk Through a Dynamic PV

2

CCE supports data redundancy policies for OBS volumes that are dynamically created.

When dynamically creating OBS volumes, you can specify a data redundancy policy (single-AZ storage or multi-AZ storage) for the volumes.

Commercial use

Using an OBS Bucket Through a Dynamic PV

3

CCE supports HostAliases for workloads.

During workload creation, you can set up HostAliases, which improve the process of mounting /etc/hosts to containers.

Commercial use

DNS Configuration Items

4

Nodes can be managed in custom node pools.

You can add a newly purchased ECS or a node that has been removed to a node pool in a cluster.

Commercial use

Accepting Nodes in a Node Pool

5

CCE supports the setting of forwarding rule priorities for ingresses.

When multiple ingresses use the same listener, sorting forwarding rules by priority can make management easier.

Commercial use

Configuring the Priorities of Forwarding Rules

6

CCE supports multiple types of nodes in yearly/monthly node pools.

Yearly/Monthly node pools support various node types to meet different application scenarios and requirements.

Commercial use

Creating a Node Pool

1

Kubernetes 1.29 clusters are in OBT.

CCE allows you to create clusters of Kubernetes 1.29.

Open beta testing

Kubernetes 1.29 Release Notes

2

CCE adds new common configuration parameters for node pools and clusters.

To make it easier to manage configuration parameters, CCE has added common parameters for in-depth configuration of core components of node pools and clusters. These parameters include the upper and lower limits for image garbage collection and container log files.

Commercial use

Modifying Cluster Configurations

Modifying Node Pool Configurations

3

Enterprise projects can be specified when CCE nodes and node pools are created.

When creating a node or node pool, you have the option to choose an enterprise project. The enterprise projects are visible in the node list, making it easier to view costs and manage resources more efficiently.

Commercial use

Creating a Node

Creating a Node Pool

4

TLS listeners are supported when Services interconnect with dedicated load balancers.

TLS can be used if ultra-high performance and large-scale TLS offloading are required. You can use TLS to forward encrypted TCP requests from clients for a Service.

Commercial use

Configuring TLS for a LoadBalancer Service

5

Advanced forwarding policies like URL redirection and rewriting are supported.

CCE LoadBalancer ingresses allow URL redirection and rewriting, as well as HTTP-to-HTTPS redirection. You have the flexibility to configure forwarding policies as needed.

Commercial use

Configuring URL Redirection for a LoadBalancer Ingress

Configuring URL Rewriting for a LoadBalancer Ingress

Redirecting HTTP to HTTPS for a LoadBalancer Ingress

6

CCE Services and ingresses that use dedicated load balancers provide an option for enabling data compression.

Services or ingresses that use HTTP/HTTPS-compliant dedicated load balancers support data compression. Enabling this feature can reduce the size of files to be transferred, improve file transfer efficiency, and minimize bandwidth usage needed for the transmission.

Commercial use

Configuring GZIP Data Compression for a LoadBalancer Service

Configuring GZIP Data Compression for a LoadBalancer Ingress

7

CCE supports General Purpose SSD V2 and Extreme SSD V2 Elastic Volume Service (EVS) disks.

CCE supports General Purpose SSD V2 and Extreme SSD V2 EVS disks. The IOPS and throughput can be customized.

Commercial use

Using an Existing EVS Disk Through a Static PV

8

CCE nodes support General Purpose SSD V2 and Extreme SSD V2 EVS disks.

When creating nodes and node pools, you can select General Purpose SSD V2 or Extreme SSD V2 EVS disks.

Commercial use

Creating a Node

9

CCE Secrets Manager for DEW (dew-provider) supports credential mounting using secrets.

With CCE Secrets Manager for DEW (dew-provider), you can easily mount credentials stored outside of clusters onto service pods using secrets.

Commercial use

CCE Secrets Manager for DEW

1

CCE Autopilot clusters are in OBT.

CCE Autopilot allows you to create serverless clusters that offer optimized Kubernetes compatibility and free you from O&M. After a CCE Autopilot cluster is created, you can deploy applications without purchasing nodes or maintaining the deployment, management, and security of nodes. You only need to focus on the implementation of application service logic, which greatly reduces your O&M costs and improves the reliability and scalability of applications.

Open beta testing

What Is a CCE Autopilot Cluster?

2

Kubernetes 1.28 now available in CCE.

CCE allows you to create clusters of Kubernetes 1.28.

Commercial use

Kubernetes 1.28 Release Notes

3

NGINX Ingress Controller metrics can be monitored.

CCE allows you to use the Cloud Native Cluster Monitoring add-on to gather metrics on the NGINX Ingress Controller. This allows you to monitor the performance and running status of the NGINX Ingress Controller in real time.

Commercial use

Monitoring Metrics of NGINX Ingress Controller

4

Multiple NGINX ingress controllers can be installed in the same cluster.

You can deploy multiple independent NGINX ingress controllers in a cluster. This allows you to customize each controller to meet specific traffic and security needs.

Commercial use

Nginx Ingress Controller

5

ELB blocklists or trustlists are supported for access control.

When creating a Service or ingress, you can add IP addresses to a trustlist or blocklist to control access to a listener of a load balancer to improve security.

Commercial use

Configuring a Blocklist/Trustlist Access Policy for a Service

Configuring a Blocklist/Trustlist Access Policy for a LoadBalancer Ingress

1

Kubernetes 1.28 clusters are in OBT.

CCE allows you to create clusters of Kubernetes 1.28.

Open beta testing

Kubernetes 1.28 Release Notes

2

Volcano supports load-aware scheduling for nodes.

Volcano Scheduler offers CPU and memory load-aware scheduling for pods and preferentially schedules pods to the node with the lightest load to balance node loads. This prevents an application or node failure due to heavy loads on a single node.

Commercial use

Load-aware Scheduling

3

Multiple subnet CIDR blocks can be configured for a node pool.

When creating a node pool, you have the option to choose multiple subnet CIDR blocks for more flexible allocation of network resources.

Commercial use

Creating a Node Pool

4

Volcano supports custom workload scale-in priorities.

After the application scaling priority policy is enabled, workload pods can be scaled in or out based on the node priority configured by users. For example, an application pod that runs on a pay-per-use node may be given priority when scaling down.

Commercial use

Application Scaling Priority Policies

5

Ingresses that use dedicated load balancers can interconnect with gRPC backend services.

Ingresses can interconnect with gRPC backend services when dedicated load balancers are used. gRPC is a high-performance general RPC open-source software framework that helps load balancers run over HTTP/2.

Commercial use

Interconnecting LoadBalancer Ingresses with gRPC Backend Services

1

Volcano supports node load-aware rescheduling.

Volcano scheduler can evict pods that do not comply with the configured policy so that pods can be rescheduled. In this way, the cluster load is balanced and resource fragmentation is minimized.

Commercial use

Descheduler

2

Volcano supports node pool affinity scheduling.

Soft affinity scheduling of a node pool is implemented based on labels in the node pool. Each node in the node pool is scored to select the optimal one for pod scheduling.

The rule is to schedule pods to nodes with specified labels as far as possible.

Commercial use

Node Pool Affinity

3

Kubernetes 1.27 now available in CCE.

CCE allows you to create clusters of Kubernetes 1.27.

Commercial use

Kubernetes Release Notes

1

CCE clusters support the dedicated load balancers that use elastic specifications.

Dedicated load balancers are suitable for fluctuating traffic and billed based on the Load Balancer Capacity Units (LCUs) used per hour.

Commercial use

Billing (Dedicated Load Balancers)

2

Custom tags can be added to dynamically created EVS disks.

When dynamically creating EVS PVCs, you can add TMS tags to automatically created EVS disks to classify resources. This facilitates cost management.

Commercial use

Using an EVS Disk Through a Dynamic PV

3

The timeout interval can be configured for a load balancer.

When a LoadBalancer Service or an ELB Ingress is used, the timeout interval can be configured.

Commercial use

Configuring Timeout for a LoadBalancer Service

Configuring Timeout for an ELB Ingress

4

Kubernetes 1.27 OBT

CCE allows you to create clusters of Kubernetes 1.27.

Open beta testing

Kubernetes Release Notes

1

Kubernetes 1.25 now available

CCE allows creating clusters of Kubernetes 1.25.

Commercial use

Buying a CCE Cluster

2

CCE Turbo ENIs support fixed IP addresses.

CCE Turbo clusters support fixed IP addresses for pods (ENIs) of StatefulSets. This is a common practice in access control, service registration, service discovery, and log audit.

Commercial use

Fixed Pod IP Address

3

CCE Turbo ENIs support automatic creation and binding of EIPs.

CCE Turbo clusters can automatically bind an EIP to a StatefulSet. This function works with the automated EIP creation function to assign a fixed EIP (fixed public IP address) to each pod.

Commercial use

Fixed Pod EIP

4

Enhances hybrid deployment of CCE Turbo clusters: Pod network priority restriction is supported.

CCE Turbo clusters balance the egress network bandwidth used by online and offline services to ensure sufficient network bandwidth for online services. When the threshold is reached for online services, the bandwidth usage of offline services will be reduced.

Commercial use

Egress Network Bandwidth Guarantee

5

CCE Turbo clusters support the association between namespaces and container CIDR blocks.

You can set container subnets and security groups by namespace to isolate network.

Commercial use

NetworkAttachmentDefinition

6

CCE Turbo clusters support CPU Burst to prevent CPU traffic limiting from affecting latency-sensitive services.

If a CPU limit is set for a container in the pod, the CPU usage of the container cannot exceed the limit. Frequent CPU traffic limiting affects service performance and increases the long-tail response latency, especially for latency-sensitive services.

CPU Burst is an elastic traffic limiting mechanism that allows temporarily exceeding the CPU limit to reduce the long-tail response time of services.

Commercial use

CPU Burst

1

Kubernetes 1.25 OBT

CCE allows creating clusters of Kubernetes 1.25.

Open beta testing

CCE Kubernetes 1.25 Release Notes

2

Huawei Cloud EulerOS 2.0 is supported.

Nodes running Huawei Cloud EulerOS 2.0 can be created in clusters of v1.23 or later.

Commercial use

Node OS

3

Enhanced CPU core binding management policies are supported.

Based on the Kubernetes static policy, the enhanced CPU core binding management policy (enhanced-static) supports burstable pods (whose requests and limits of CPU are positive integers) and allows certain CPUs to prioritize these pods, enhancing application stability.

Commercial use

Enhanced CPU Core Binding Management Policy

4

CCE clusters support monitoring components of master nodes.

CCE clusters can monitor the kube-apiserver, kube-controller, kube-scheduler and etcd-server components of master nodes.

Commercial use

Monitoring Metrics of the Master Node Component

1

Workload identity

With workload identities, your workloads in a cluster can access cloud services like IAM users without using the AK/SK, reducing security risks.

Commercial use

Workload Identities

2

Resource reservation model V2

For clusters of v1.21.4-r0, v1.23.3-r0, or later, the node memory reservation model is optimized to V2 and can be dynamically adjusted using the node pool parameters kube-reserved-mem and system-reserved-mem.

Commercial use

Formula for Calculating the Reserved Resources of a Node

3

Support for containerd

You can select containerd as the container runtime for CCE clusters.

Commercial use

Node Overview

4

Custom security group for a node pool

You can specify a security group when creating a node pool.

Commercial use

Creating a Node Pool

5

Enhanced npd add-on

CCE enhances the monitoring on component events, OS, network, and storage, and provides fault isolation.

Commercial use

npd

1

CCE Turbo clusters now GA

CCE Turbo clusters run on a cloud native 2.0 infrastructure that features software-hardware synergy to support passthrough networking, high security and reliability, and intelligent scheduling.

Commercial use

Buying a CCE Cluster

2

Kubernetes 1.23 now available

CCE supports creating clusters of Kubernetes 1.23.

Open beta testing

Buying a CCE Cluster

3

Kubernetes 1.21 now available

CCE supports creating clusters of Kubernetes 1.21.

Commercial use

Buying a CCE Cluster

4

Specifying enterprise projects for PVCs

By default, the enterprise project to which a PVC belongs is the same as that to which the cluster belongs. You can now specify an enterprise project.

Commercial use

PersistentVolumeClaims (PVCs)

5

Support for local PVs and ephemeral volumes

Data disks on CCE nodes can be configured as local PVs or ephemeral volumes to provide better data read and write performance.

Commercial use

Local PVs and EVs