The Problem: From Gigabytes to Terabytes Overnight
When Vanderbilt University’s Center for Structural Biology brought in its first cryo-electron microscopy (cryo-EM) instrument, data storage demands skyrocketed. What had previously been a manageable 100 GB per day with X-ray crystallography suddenly ballooned to 20 TB daily.
“It was incredible. Three orders of magnitude more than we were used to.”
This massive jump created an urgent need for scalable, high-performance storage. Existing systems weren’t built to handle this influx, and traditional storage strategies became immediately obsolete.
The Challenge: Managing Unstructured Scientific Data at Scale
Cryo-EM, along with techniques like NMR and X-ray crystallography, generates enormous volumes of unstructured data. Beyond storage capacity, the team needed reliability, high availability, metadata management, and robust backup strategies. On top of that, government mandates required data retention and public accessibility for up to 10 years, a costly, unfunded burden.
“Any data from a federally funded grant has to be retained and publicly accessible for up to 10 years. That’s an unfunded mandate, but it’s the direction we’re going.”
Additionally, researchers needed efficient ways to restore accidentally deleted files, replicate data across systems, and maintain uptime despite potential hardware failures.
The Solution: TrueNAS M-Series with ZFS Snapshot Replication
In 2019, Jarrod’s team deployed a TrueNAS M-Series appliance with over 600 TB, equipped with dual controllers and redundant HBAs. The system was built to perform large-scale ZFS snapshot replication from local Linux workstations.
“I learned how to do snapshot-based incremental backups with ZFS. We never have to walk file systems anymore. It’s an effective solution, and relatively easy to deploy and manage.”
This design met their backup and performance needs while maintaining flexibility and cost efficiency.
The Build: Leveraging ZFS for Scientific Reliability
Jarrod had long experience with enterprise file systems like GPFS but found ZFS better suited to their needs.
“Snapshots let users recover files they accidentally deleted. And for backup, you don’t have to worry about data changing out from under the software. That became a huge deal for us.”
The team took full advantage of ZFS features:
- Snapshot replication
- Automatic change detection
- Metadata acceleration with SSDs
He acknowledged that ZFS has a learning curve, but the payoff is immense for those who master it.
“ZFS gives you a lot of rope to work with, and potentially to hang yourself with. But once you learn it, it’s hard to go back.”
How TrueNAS Helped: Scalable Performance, Cost Efficiency, and Strong Support
Cost-Effective Licensing
One of the biggest wins for Vanderbilt was TrueNAS’s pricing model:
“You don’t nickel and dime us with capacity-based licensing. That is the number one thing that kills a product for us. Keeping data isn’t a profit center. It’s a liability. So taxing it is the death of a solution.”
High-Availability Hardware
When a hardware component failed, TrueNAS’s HA configuration worked flawlessly.
“One HBA failed. You walked us through it over the phone and got it swapped without downtime. The HA worked. I was super impressed.”
Expert Support
Jarrod’s technically skilled team doesn’t need hand-holding, but they value fast access to experts when necessary.
“When we call support, it’s already a complex issue and your team matches us with the right engineer quickly.”
A Word to Other Researchers
Jarrod offers a clear recommendation to peers:
“I’d recommend TrueNAS to anyone doing large-scale backup, especially in scientific computing. If you’re managing petabytes of unstructured data, this is the right tool.”
He cautions labs not to underestimate the speed at which advanced imaging can overwhelm traditional storage systems:
“Data from cryo-EM and other advanced imaging platforms can overwhelm traditional storage setups almost overnight. You need solutions that scale and that won’t lock you into expensive licensing models.”


