Operating Steps

1. Log in to the Baidu AI Cloud platform:

   • If you don't have a username, complete the registration first. For operations, see Register Baidu Account.
   • If you already have a user name, see Login for the login operation.
2. After the successful login, select "Product and Services> CloudNative> Cloud Container Engine (CCE)" and enter the "Cluster Management> Cluster List" page.
3. (Optional) The “Cluster List” page displays the information of all created clusters, such as name/ID, status, and region. The user can search clusters via the cluster name.
4. (Optional) To select a region, you can switch between regions according to the actual requirements. The Cloud Container Engine (CCE) service is available in North China-Beijing, North China-Baoding, South China-Guangzhou, East China-Suzhou, Hong Kong, and other regions.

5. Click "Create Cluster" to enter the "Select Template" interface. Then, select a template according to the template introduction and business requirements.

   • By default, the cluster quota in a single region is five, and the node quota in each cluster is twenty.
   • If required, you can submit a ticket to increase the cluster quota in a single region.

6. Click "Create" to create the navigation according to the cluster and configure it according to your business requirements, including basic configuration, network configuration, Master configuration, and Worker configuration.
7. Basic configuration

8. Network configuration

The container network is an independent address space, which must be independent of the node network, node subnets, and container networks of other clusters.

9. Master configuration

Users can choose different Master types according to their business requirements,

• Hosting Master: means that the Master node is wholly configured, created, and maintained by Baidu AI Cloud and is invisible to users. Users do not need to pay for the hosting Master. But in this mode, there are a minimum number of Workers in the cluster.

  

• Custom Master: Users can customize the number of Master replicas and machine configurations according to business scale and requirements. The configuration method is the same as "Worker Configuration" below). There is no limitation on the minimum number of workers in the custom master cluster.

  

10. Worker Configuration

Click “Add configuration” to pop up the "Add node" dialog box. Users can select "Create node”. However, the new node is available only in post-paid mode. If you want to adopt the prepaid method, you can first create a prepaid node on the BCC page and select "Existing node" and then the "Created prepaid node".

Model

You select different Baidu Cloud Compute types based on your business requirements, e.g., selecting Ordinary type. Configuration details are shown in the following figure:

GPU Configuration Instructions: The operating system supports the integrated GPU driver system configured by Baidu AI Cloud. If you create GPU custom images in BCC, the images are GPU custom images. If you need to use custom images, set up the integrated GPU driver manually. GPU provides the local disk with a fixed size and doesn’t support independent choice temporarily. Information is shown in the following table:

Type of GPU Card	Quantity of GPUs (block)	GPU (core)	Memory (GB)	Disks (GB)
NVIDIA Tesla K40	1	6	40	200
NVIDIA Tesla K40	2	12	80	400
NVIDIA Tesla K40	4	24	160	800
NVIDIA deep learning development card	1	6	40	200
NVIDIA deep learning development card	2	12	80	400
NVIDIA deep learning development card	4	24	160	800

Storage

• Users can select Create CDS Disk or Create CDS Disk from Snapshot. How to select SSD Cloud Disk/High Performance Cloud Disk/Standard Cloud Disk，view Detailed Description

• At present, the system can achieve the automatic formatting of the created CDS Cloud Disk Storage, the mounting of Cloud Disk Storage to the directory/data. Also, the system designates the storage target of the container as /data/docker. This feature only supports the first CDS Cloud Disk Storage.

  

Security Group

The CCE’s default security group/custom security group. You can change access rules after creating instances, see Manage Security Group

Public IP

• If you select “Purchase Elastic IP”, you need to configure EIP. For details, see EIP Instance.
• In Public Bandwidth, either Bill-by-Traffic or Bill-by-Bandwidth is used as the billing standard.

11. After the user confirms the selected configuration information and configuration cost, click "Finish" to start creating the cluster.
12. Click “Management Console”, enter the “Cluster List” interface, and view the status of the created clusters. After you create the cluster successfully, the cluster status displayed is “Running”.