Backup is Broken

May 26, 2021 | Andres Rodriguez Backup is Broken

File backup is broken and no one is talking about it. The most recent uptick in ransomware attacks has exposed a fallacy that the industry has been supporting and re-enforcing for decades. Large companies running modern backup software and following recommended best practices are being brought down to their knees for weeks by ransomware attacks. How is this happening?

All backup technology relies on a simple premise. You make a copy of the data in order to protect it. The restore phase — the reason we rely on backup in the first place — moves that copy back into production. If the production system is a VM or a database, then it doesn’t take that long to copy back the healthy data in order to recover. However, when we are talking about even a moderately sized, multi-terabyte file server, the restore operation can take hours or even days.

This limitation has nothing to do with any specific backup vendor technology. The problem is inherent to the entire approach, and the backup copy-back operation which is necessary in order to restore a file server. It takes a long time to rebuild a file server because — as anyone who has ever had to wait for a migration job to complete knows — it takes a long time to copy files to any file server.

The more files to copy, the more time you wait. Multiply that by the many file servers in a large organization that could be compromised by a full-blown, distributed ransomware attack. Now you are looking at a multi-week recovery effort that is pure IT Hell.

Backup protects your files, but it does not protect your business.

The alternative to file server backup is a file system versioning technology. Unlike backup, a versioned file system can use internal pointers to restore files back to previous, healthy versions of themselves. No data has to be copied back; hence, the restore operation is nearly instantaneous, no matter how many files need to be restored.

The problem with traditional, block-based, versioned file systems is that the number of versions are limited. This means that traditional versioning, or snapshots, can typically be used to unwind damage that has occurred only a few days back.

Hackers know this weak spot. So they wait. A well-executed ransomware attack is not unlike a WWII Wolfpack offensive. U-boats would lay under the surface, undetected, for days or weeks, tracking and surrounding Allied convoys before attacking. Similarly, hackers know that if their ransomware goes undetected for just a few days, operators will be forced to rely on backups in order to restore their file servers. They know that this will, in turn, extend the recovery window to such a point that it becomes smarter for the company to pay the ransom than to wait for restores to complete. The files are still protected by the backup, but they are useless. The business is out of time.

This is the hidden flaw in backup that hackers have become so adept at exploiting. Today’s ransomware attacks target dozens of file servers simultaneously, covering hundreds of terabytes, and go undetected for weeks. A modern, cloud-based versioned file system like UniFS® can recover from large-scale ransomware attacks without having to resort to backup, regardless of how long the attack goes undetected. Unlike traditional file systems, cloud-based systems have no limitations on the number of versions or how far back in time these versions can exist. Because the versions never need to be re-used, they are immutable, meaning no malicious attack can ever change the past. A cloud-based versioned file system is forever immutable and immediately recoverable. We know it works. Our customers remind us all the time.

Backup, on the other hand, doesn’t protect you against the long outage that results when you try to restore those files. It is a bad insurance policy against ransomware. At our user groups, we ask our customers to name their favorite thing about Nasuni. Again and again, the answer at the top the list is some variation of not having to do backups or not having to rely on backup or not having to wait for backups.

Backup is broken. Backing up files is bad for business. It is a Wile E. Coyote insurance policy. You feel fantastic, with your business running at full speed, until ransomware strikes and you realize you are way out there past the edge of the cliff, standing on air and spinning your legs.

Nasuni protects your files and your business. We can recover your files without backups, without delays and without giving hackers any leverage. We turn what might have been your worst day into an IT win and internal success story. And we do all of this at scale.

No surprises. No lies. That’s a real promise.

Related Posts

June 07, 2022 VMware Highlights Nasuni as Horizon File Data Solution

This has been an exciting year for Nasuni on several fronts. We’ve signed on some of the largest companies in the world as new customers. Our existing customers are expanding their use of the platform. The investment community has given us another huge vote of confidence, our rapid ransomware recovery capabilities are pulling in unprecedented business, and our engineering team continues adding new features to the platform. As the VP of Strategic Alliances, though, I’m particularly excited about how our partnerships with some of the top technology providers in the world continue to get stronger.

, , , , ,
May 24, 2022 The Sanctity of Data: Why I Joined Nasuni

If the energy industry teaches IT professionals one thing, it’s the importance of respecting the sanctity of data. Whether you devote internal resources to acquire seismic data or purchase it from another organization, this data is gold. Highly paid geoscientists and engineers need to be able to analyze it quickly and securely, wherever they might be in the world. If this data is lost, corrupted, or incomplete, these professionals can’t do their jobs, and the company loses money.

, , , , ,
May 17, 2022 One Global Customer’s Cloud Journey

Our customers typically come to us with an eye on solving one or two major, pressing problems. Maybe a StorSimple array is approaching EOL or they need to exit a handful of data centers. A ransomware incident may have sparked their interest. Or it could be that the high costs and extreme hassle of managing backup across multiple locations sent them searching for a more efficient cloud alternative.

, , , , ,