Storage and Research Data on Sagehen HPC

Where your data lives on the cluster, how much space you have, and how to keep sensitive data safe. Two minutes here can save you a lost dataset later.

Four places your data can live

Two persist between sessions; two exist only while a job runs.

/rhome

Your home directory: scripts, code, configuration, small files.

✓ PERSISTS

/bigdata

Your lab’s shared research data on the BeeGFS parallel filesystem.

✓ PERSISTS

/scratch

Fast local SSD for intermediate files while a job is running.

⚠ DELETED WHEN THE JOB ENDS

/tmpfs

RAM-backed space for the fastest possible I/O during a job.

⚠ DELETED WHEN THE JOB ENDS

The one rule that saves datasets: anything in /scratch or /tmpfs disappears the moment your job completes. Not after 90 days; at job completion. Make copying results back to /bigdata the last line of every job script.

1 TB
per lab

Your quota

/rhome and /bigdata share a 1 TB quota per lab. Check your usage any time with the quota_check.sh command. Need more? Additional storage can be purchased; talk to us about options and pricing.

Know your data’s tier

Every dataset on Sagehen falls into one of three tiers, which determine how it is stored and who can see it.

PUBLIC

Data that can be shared openly: published datasets, open source code, public reference data.

PROPRIETARY

Unpublished research data visible to your lab only. Typical for most work in progress.

RESTRICTED

Sensitive data (human subjects, contractual restrictions) stored in encrypted form (AES‑256). Set up with us before the data arrives.

Moving data in and out

Drag and drop through the OnDemand file browser, use scp/rsync from a terminal, or connect cloud storage with rclone. Our Data Transfer workshop covers all three.

Planning a data-heavy project?

Before you rely on the cluster as the only home for irreplaceable data, talk to us about what is backed up, what is not, and the right plan for your project. [CONFIRM backup statement with team before publishing]

Sensitive data? Big data? Just unsure?

A short conversation before data arrives beats untangling it afterward. We’re happy to help you plan.

its-hpc@pomona.edu

Last reviewed: August 2026