When I try to start my Amazon EMR cluster, the cluster fails to start, and I get the error message "application provisioning failed."

EMR clusters support multiple configuration options, which are usually deployed through bootstrap actions. When you launch a cluster, Amazon EMR runs bootstrap actions after provisioning the Amazon Elastic Compute Cloud (Amazon EC2) instances. After bootstrap actions are completed, Amazon EMR installs the native applications that you specify and makes sure that the cluster is properly configured. For more information, see Understanding the Cluster Lifecycle.

If the cluster can't perform these steps, the cluster fails to start, and you get the "application provisioning failed" error message. The most common reason for the problem is the cluster configuration. For example, an EMR cluster fails to start if incorrect credentials are specified for an external Hive metastore.

Review your cluster’s node provisioner logs to determine the cause of the error. The logs are located in a folder on Amazon Simple Storage Service (Amazon S3):

s3://<logging-bucket>/<cluster-id>/node/<instance-id>/provision-node/apps-phase/

There are two log files in this folder:

  • stderr.gz: Refer to this file to see the main error message.
  • stdout.gz: Refer to this file for more details about the error, such as the stack trace.

Did this page help you? Yes | No

Back to the AWS Support Knowledge Center

Need help? Visit the AWS Support Center

Published: 2016-05-13

Updated: 2018-11-14