Morgridge Datavault Explainer

As a lab’s data stores grow, the need to manage them becomes more urgent. Knowing where the data are stored, who they belong to, how long they need to be kept, what grants they might be associated with, and who has permission to access them becomes a stiff challenge when the data are unorganized.

The Datavault is a model that DiscoverIT uses to help labs organize their data and prepare for the questions that might be asked over the lifetime of the data. Rather than just purchasing storage servers and storing data there in one large pool, the model divides data up into smaller chunks that have their own access controls, archival policies, and quotas. DiscoverIT can monitor the data stored this way and help labs to manage their data stores more effectively.

In practice, the Datavault is just a system of quotas and documentation. When a new storage pool is required, a designated lab member will fill out a form requesting the space and DiscoverIT will provision a new project under the Datavault with a size quota, access list, and proposed archival date based on the data in the form.

Organizing Your Data

Choosing how to organize the data can be difficult but one of the benefits of the Datavault model is flexibility, allowing labs to organize their data in a way that fits their workflow. One path is to organize data based on the projects the data are being used for. Alternatively, a lab could organize data based on the people working on it, with a shared pool for things everyone is working on.

A lab could also use a mix of these models, having several separate projects for specific data alongside pools devoted to personal collaborations, or any other setup that fits the data that lab has. The Datavault setup will provide some clarity about what data is where and who is supposed to access it, and it gives DiscoverIT the capacity to help a lab manage and plan their data use over time.

Data Migration Process

To migrate data into the Datavault model, we’ll follow this process:

1. DiscoverIT will meet with the PI and any person from their lab they might designate to head up the migration process to get a high-level understanding of what and where the current data are, and what should be moved to the Datavault.
2. DiscoverIT will work more closely with a designated lab member to develop a plan for creating projects within the Datavault and scheduling migration of data into those projects.
3. As projects get scheduled, DiscoverIT will make sure that we have access to do the migration and perform the migration when it’s ready.
4. If some data is getting archived instead of migrated, DiscoverIT will begin that process, including creating a complete list of the files getting archived and verification of the file contents.