E-MapReduce:Create a gateway cluster

Prerequisites

Before you begin, ensure that you have:

• An existing Hadoop or Kafka cluster in EMR with a Running status. For setup instructions, see Create a cluster.

Limitations

This procedure applies only to Hadoop and Kafka clusters. For gateway environment deployment on DataLake, OLAP, DataFlow, and Custom clusters, see Gateway deployment modes and selection guide.

Create a gateway cluster

1. Log on to the EMR console.

2. On the EMR on ECS page, click the name of the target cluster.

3. In the upper-right corner of the Basic Information page, choose All Operations > Create Gateway.

4. On the Create Gateway page, configure the parameters described in the following table.

   Associated settings

   Parameter	Description
   Region	The physical location of the gateway cluster.
   Resource Group	The resource group for the gateway cluster. To create a new resource group, click Create Resource Group. For more information, see Create a resource group.
   Associated Cluster	The compute cluster to associate with the gateway cluster, filtered by the selected region. The cluster must be in Running status and must be a Hadoop or Kafka cluster. After you select a cluster, the gateway cluster's VPC defaults to the VPC of the associated cluster. Clusters created in both the new and old console versions are supported.

   Basic settings

   Parameter	Description
   Billing Method	The billing method for the gateway cluster. Subscription: Pay before use. Pay-as-you-go: Charged hourly based on actual usage. Suitable for short-term tests or dynamic workloads.
   Zone	The zone where the associated cluster is located.
   vSwitch	The vSwitch in the corresponding VPC and zone.
   Default Security Group	The security group of the associated cluster.
   Assign Public Network IP	Whether to attach an Elastic IP Address (EIP) to the gateway.
   Node Group	The compute resources for gateway nodes. Configure the following sub-parameters: Instance Type: The ECS instance types available in the selected region. For more information, see Instance families. System Disk: The system disk type. Supported types: ultra disk, enterprise SSD (ESSD), and standard SSD. Available types vary by instance type and region. Size range: 60 GiB to 500 GiB. Released with the cluster by default. Data Disk: The data disk type. Supported types: ultra disk, ESSD, and standard SSD. Available types vary by instance type and region. Size range: 40 GiB to 32,768 GiB. Released with the cluster by default. Instances: The number of gateway nodes. Default: 1.
   Cluster Name	The name of the gateway cluster. Must be 1–64 characters and can contain Chinese characters, letters, digits, hyphens (-), and underscores (_).
   Identity Credentials	The credentials used to log on to all nodes in the gateway cluster. Password: Must be 8–30 characters and include uppercase letters, lowercase letters, digits, and special characters (!@#$%^&*). Key Pair: Select a key pair for SSH access. If you haven't created one, click Create Key Pair to go to the ECS console. Keep the private key file (.pem) secure — it is required each time you log on via SSH.

   Advanced settings

   Parameter	Description
   ECS Application Role	The Resource Access Management (RAM) role that grants applications on the cluster permission to call other Alibaba Cloud services. Default: AliyunECSInstanceForEMRRole.
   Bootstrap Actions	Optional. Custom scripts that run before the cluster starts. For more information, see Run scripts using bootstrap actions and Manually run scripts.
   Tags	Optional. Tags to attach to the cluster. Tags can be added at creation time or after. For more information, see Manage tags.
   Data Disk Encryption	Optional. Encrypts the data disk. This setting can only be enabled at cluster creation time. For more information, see Enable data disk encryption.

5. Click Create and Pay. The cluster Status changes from Creating to Running when the gateway cluster is ready.

What's next

After the gateway cluster status changes to Running, connect to it via SSH using the credentials you configured during creation. Once connected, use the cluster's client tools to submit jobs to the associated Hadoop or Kafka cluster.