Files and Volumes

Two distinct concepts

Type	Description	Typical use‑case
User Volumes	Private, per‑user directories that you can create, mount, and share. Persist across container restarts.	Your notebooks, scripts, intermediate results, or any data you own.
Data Volumes	Curated, read‑only (or read‑write if granted) datasets administered by platform admins and shared across projects.	Large scientific data sets (catalogs, calibration files, reference models) that many users need.

Why the distinction matters

User volumes live in your personal namespace (/workspace/Storage/<username>/…). You control who can see them.
Data volumes are centrally maintained; admins decide who can mount them (/workspace/<datavolume_name>/). This prevents accidental modification of shared data.

How to work with them

Browse the Files web‑app

Clicking on Files on the main dashboard you can see both User Volumes and Data Volumes on the left.

User Volumes : Should have at least 2 folders, persistent and scratch. Mapped to Storage and Temporary on your compute containers and compute jobs. Shared folders from other users with you will appear at this level as well.
Data Volumes might appear empty, until the relvent permissions are given to you. Make sure to reach out to your supervisor or MPE point of contact to request access to the Data you need.

Create a new Volume (optional)

You can do it from the UI (Create Volume button) and select the root volume (Storage or Temporary):

Suppose Storage is chosen, then the UI will automatically mount the new volume at /workspace/Storage/$SCISERVER_USER_NAME/my_volume.

Mount volumes when you launch a Compute container

In the Create Container dialog:

Section	What to select
User Volumes	Choose one or more of your private volumes.
Data Volumes	Pick any shared datasets you need for the analysis.

They are mounted inside the container as:

User volumes → /workspace/Storage/$SCISERVER_USER_NAME/<volume>
User volumes → /workspace/Temporary/$SCISERVER_USER_NAME/<volume>
Data volumes → /workspace/<data_volume_name>/

Quota limits

Although we don’t enforce a specific quota. We do monitor for fair-use. What this means is that Sciserver is a limited and shared resource, and so we aim to keep a balance between scientific needs and remaining storage. If you need exceptional amounts of storage for your project, this should be previously discussed with your MPE point of contact and agreed upon.

Even though we attempt to monitor this use, we kindly ask users to be proactive and reach out if you feel your needs exceed more than 1 Terabyte in size.