EFS Datasets

Last updated 10 months ago

This feature is only available on some Clusterone Enterprise installations. Contact your account administrator for details.

Jobs on Clusterone can use data stored on Amazon's Elastic File System (EFS) service.

Your EFS drives are integrated into your Clusterone Enterprise setup upon installation. After the installation is complete, your EFS drives are automatically mounted to each pod created by Clusterone.

Runtime

If you have an EFS dataset, it will become available in /data/ at runtime by default. There is no need to add the dataset to the job specification.

As EFS datasets are mounted on the pods, data is not copied from the dataset to the pod at runtime. Changes made to /data/ are transparently made to the data.

Note that modifying an EFS file while running a job reading it may result in the job to fail. Make sure no other user is modifying a file if you use it for a job.

This page describes the default configuration, please contact your administrator for specifics on how data is managed in your Clusterone Enterprise installation.