Restart EMR Services with Zero Downtime via Rolling Mode - E-MapReduce

Restart a service after modifying its configuration to apply the changes, or when a service is faulty or behaving abnormally.

Prerequisites

Before you begin, ensure that you have:

Use rolling mode to minimize impact on running jobs. For services in primary/secondary mode, restart the secondary instance first, then the primary instance.
Restarting a service is not the same as restarting a cluster. EMR does not support cluster-level restart.
To stop a restart in progress: in the upper-right corner of the Services tab, click Operation History. In the Operation History panel, find the operation and click Terminate in the Actions column.
Important
Terminating a restart may cause configuration inconsistency across nodes in a node group.

Go to the Services tab.
1. Log on to the EMR console. In the left-side navigation pane, click EMR on ECS.
2. In the top navigation bar, select the region where your cluster resides and select a resource group.
3. On the EMR on ECS page, find the cluster and click Services in the Actions column.
On the Services tab, find the service to restart, move the pointer over the icon, and select Restart. Alternatively, go to the Configure tab of the service. In the upper-right corner, choose More > Restart.

In the dialog box, configure the parameters and click OK.

Parameter	Description
Rolling Execution	Enabled by default. Runs tasks one node at a time at the interval set in Rolling Execution Interval. Important Disabling this option restarts the service on all nodes simultaneously, making the service unavailable during the restart.
Rolling Execution Interval	The interval between two consecutive task executions during a rolling restart.
Failure Handling Policy	The action to take when a node fails its health check during the restart: Proceed Upon Failure of Single Node: Skip the failed node and continue on remaining healthy nodes. Stop Upon Failure of Single Node: Stop the entire restart operation.
Execution Reason	The reason for restarting the service.

In the Confirm dialog box, click OK. The system displays The operation is successful. and runs the restart according to the policy you configured.

To check whether the service and its components are healthy after the restart, see View the health status of services and components.
For common service issues, see FAQ about cluster management.