Monitoring an execution - AWS Elastic Disaster Recovery
Services or capabilities described in AWS documentation might vary by Region. To see the differences applicable to the AWS European Sovereign Cloud Region, see the AWS European Sovereign Cloud User Guide.

Monitoring an execution

To follow an execution in the console, open the plan and choose the execution from its execution history. The execution details show the status of the execution, the status of each step, and the status of each server within a step, including the recovery job that AWS Elastic Disaster Recovery created for it.

With the AWS CLI, use get-recovery-plan-execution for the overall status and list-recovery-plan-execution-steps for per-step progress. Use list-recovery-plan-executions to see the history for a plan, optionally filtered by status.

aws drs get-recovery-plan-execution \
    --recovery-plan-execution-arn EXECUTION_ARN

aws drs list-recovery-plan-execution-steps \
    --recovery-plan-execution-arn EXECUTION_ARN

An execution reports one of the following statuses:

Execution status Description

CREATED

The execution was accepted and its source servers are being validated. No servers have been recovered yet.

IN_PROGRESS

AWS Elastic Disaster Recovery is running the steps of the plan.

COMPLETED

Every step reached a terminal state without a critical failure. An execution that skipped one or more steps also reports COMPLETED.

FAILED

Either the validation at the start of the execution failed, or a step failed because a critical server did not recover. The remaining steps did not run. If a step failed, you can retry or skip that step to continue. If the validation failed, no step ran, so there is nothing to retry or skip; correct the problem the error reports and start a new execution.

TIMED_OUT

The execution reached its 24-hour limit while a step was still running. The remaining steps did not run, and you cannot retry, skip, or resume the execution. Start a new execution of the plan to finish the recovery.

CANCELLING

You canceled the execution and AWS Elastic Disaster Recovery is waiting for the step that was already running to finish.

CANCELLED

The execution stopped after you canceled it. The steps that had not started were canceled.
Note

If an execution completes with skipped steps, the status is still exactly COMPLETED. There is no separate status value for it. The detail appears as a message on the execution: Completed with skipped steps. View execution details for more information. If you script against these APIs, branch on the status field and treat the execution's errorDetail as informational text only. In particular, do not branch on errorDetail.code, which is not a reliable signal of failure — a successful execution that skipped a step still returns a code value.

Each step within an execution reports one of the following statuses:

Step status Description

NOT_STARTED

The step is waiting for the steps before it to finish.

EXECUTING

A server step is recovering its servers.

WAITING

A wait step is counting down its wait duration.

COMPLETED

The step finished. A server step can complete even if optional servers failed.

FAILED

At least one critical server in the step did not recover.

TIMED_OUT

The step was still running when the execution reached its 24-hour limit.

SKIPPED

You skipped the step, or the step was not run because the execution was canceled.

Recovery plan API calls are recorded by AWS CloudTrail, so you can audit who created, changed, or ran a plan. For more information, see Logging AWS Elastic Disaster Recovery API calls using AWS CloudTrail.