Overview

Overview

Log in to MRS Manager, click Cluster, click the name of the desired cluster, and choose Service > KrbServer. On the displayed page, click Instance. The displayed instance management page contains the function area and role instance list.

Functional Area

After selecting the instances to be operated in the function area, you can maintain and manage the role instances, such as starting or stopping the instances. Table 1 shows the main operations.

Table 1 Instance maintenance and management

UI Portal

Description

Start Instance

Start a specified instance in the cluster. You can start a role instance in the Not Started, Stop Failed, or Startup Failed state to use the role instance.

More > Stop Instance

Stop a specified instance in the cluster. You can stop a role instance that is no longer used or is abnormal.

More > Restart Instance

Restart a specified instance in the cluster. You can restart an abnormal role instance to restore it.

More > Instance Rolling Restart

Restart a specified instance in the cluster without interrupting services. For details about the parameter settings, see Performing a Rolling Restart of a Cluster.

More > Decommission/Recommission

Recommission or decommission a specified instance in the cluster to change the service availability status of the service. For details, see Decommissioning and Recommissioning an Instance.

Note

Only the role DataNode in HDFS, role NodeManager in Yarn, and role RegionServer in HBase support the recommissioning and decommissioning functions.

Desired instance > More > Synchronize Configuration

If the Configuration Status of a role instance is Expired, the role instance has not been restarted after the configuration is modified, and the new configuration is saved only on MRS Manager. In this case, use this function to deliver the new configuration to the specified instance.

Note

  • After synchronizing the role instance configuration, you need to restart the role instance whose configuration has expired. The role instance is unavailable during the restart.

  • After the synchronization is complete, restart the instance for the configuration to take effect.

Desired instance > Instance Configurations

For details, see Managing Instance Configurations.

You can filter instances based on the role they belong to or their running status in the function area.

Note

Click Advanced Search to search for specified instances by specifying other filter criteria, such as Host Name, Management IP Address, Business IP Address, or Instance Groups.

Role Instance List

The role instance list contains the instances of all roles in the cluster. The list displays the running status, configuration status, hosts, and related IP addresses of each instance.

Table 2 Instance running status

Status

Description

Normal

Indicates that the instance is running properly.

Faulty

Indicates that the instance cannot run properly.

Decommissioned

Indicates that the instance is out of service.

Not started

Indicates that the instance is stopped.

Unknown

Indicates that the initial status of the instance cannot be detected.

Starting

Indicates that the instance is being started.

Stopping

Indicates that the instance is being stopped.

Restoring

Indicates that an exception may occur in the instance and the instance is being automatically rectified.

Decommissioning

Indicates that the instance is being decommissioned.

Recommissioning

Indicates that the instance is being recommissioned.

Failed to start

Indicates that the service fails to be started.

Failed to stop

Indicates that the service fails to be stopped.

Instance Details

You can click an instance name to go to the instance details page and view the basic information, configuration file, instance logs, and monitoring metric reports of the instance.