百度智能云

All Product Document

          MapReduce

          Getting Started

          1. Initialize a BmrClient.

            BmrClient is a client interacting with BMR services and used by BMR to complete all operations. You can initialize the client by referring to create BmrClient).

          2. Create a BMR Cluster (cluster).

            Create a BMR cluster to submit and run the specified step. Create a cluster before submitting the step. After the request for creating clusters, the new cluster's ID is returned and used for subsequent operations.

            When creating clusters, you need to specify image type, image version number, and virtual machine instance group configuration for the virtual machine instance. The configuration includes cluster name, information on the component to install (e.g., Hive, Pig, and HBase), BOS path used to upload the operating log, and information on the step to operate. You can perform the cluster configuration by referring to create cluster).

          3. View the information on cluster status.

            You can view the information on cluster status through the cluster ID. The cluster becomes available after a certain period, and after then, the cluster begins to call and perform the steps submitted to it.

          4. View the information on the cluster instance group.

            You can view the information on the cluster instance group (including instance group type, number of instances in the instance group, etc.) through the cluster ID.

          5. Modify instance group configuration.

            You can modify the configuration of the CORE or TASK instance group through cluster ID and instance group ID.

          6. View the information on the instance.

            You can view the information on an instance through cluster ID and instance group ID.

          7. Add one or more Steps. Add one or more step tasks to the specified cluster.

            After creating the cluster, you can add different types of steps (including Custom Jar, Streaming, Hive, and Pig) according to applications installed on the cluster. After the request for adding step, an array including IDs of all new steps is returned, and the step ID is used for subsequent operations.

          8. Check the information on the step's operating status.

            You can check the information on the step's operating status through the step ID.

          9. Terminate the cluster.

            After completing the step, you can send the request for terminating the cluster through the cluster ID. The termination of the cluster releases the cluster's virtual machine instances and stops the billing.

          Previous
          Introduction
          Next
          Install SDK Kit