Protecting Unstructured Data: An Exercise in Determination

Whether on-premises or in the cloud, unstructured data continues to increase in business criticality, yet remains woefully neglected.


As the very public war rages on in our battle against ransomware and other data loss vulnerabilities, it is refreshing to read that some companies are making headway, as this recent report from Databarracks suggests.

Unfortunately, however, it appears that while structured data appears to be a bit safer of late, unstructured data still seems to be lagging behind. Whether on-premises or in the cloud, housed in object or network-attached storage (NAS), unstructured data continues to increase in business criticality, yet remains woefully neglected.

This could be due, in part, to the fact that protecting unstructured NAS and object data presents some particular challenges. As a result, while most companies and vendors have focused on protecting databases and virtual machines (VMs), business continuity planning and protection for these other vital assets remain something of a poor relation, at increasing risk of disruption and loss.

When NAS or object storage has been targeted by hackers, most organizations will have backups, but will often struggle to restore them if their entire system is down and they don’t have access to identical hardware on which to quickly rebuild the system.  In addition, there is the issue of hackers attacking the backup first, before moving on to the production data.

In either of these situations, the first hurdle to overcome is that of how long it will take to regain access to the most essential elements of unstructured data needed to keep the organization going. It’s only at this point that too many IT teams embark on a considered analysis of their situation to identify what elements of their unstructured data are the ‘must-haves’ and how their data protection strategy must change for the better to prevent similar problems in the future. (Some call this the “Too Little, Too Late” or TLTL strategy.)

However, if TLTL is not your ideal approach, the first step should always be to make a ‘golden copy’ of your most critical unstructured datasets and keep them offsite (data center or cloud) in their native, easily accessible format, and behind a firewall where just a few trusted people in the organization are allowed access. And critically, no trace of where this copy lives should be left on the production or any other system giving hackers clues to where they can find your golden copy.

In a cybersecurity context, this approach arms organizations with the tools to frustrate even the most tenacious of criminals in their attempts to find and hijack your operations’ critical unstructured data. And then, if you need to open access your golden copy to recover lost data, the process should be to immediately and automatically set up another golden copy staying one step ahead of cyberthreats.

The difference between organizations that follow this path of determination and those who choose to risk their unstructured data is dramatic. The latter face the prospect of uncertainty, downtime, business disruption, and even business failure (and possibly career disruption for the unlucky individual(s) held responsible). The former place themselves in a position of power and even competitive advantage, knowing that their unstructured data can be recovered in the face of virtually any scenario and that they can get back to business as usual, with barely a blip on the radar.


Michael Jack has more than 30 years of experience delivering complex IT solutions to companies big and small, with the last 15 years in the data storage industry. During his tenure at EMC, Jack helped to develop Centera into one of the highest quality products in the EMC portfolio. Jack co-founded Datadobi and has helped build it into a highly respected company focusing on unstructured data management.