Your browser is out-of-date!

Update your browser to view this website correctly. Update my browser now


Enterprise Storage: Affordable Reliability for Government Video Operations

A commentary by NetApp’s Jason Danielson.

Hundreds of applications from video editing, to media asset management, to video distribution and video surveillance are integrated with IT storage today.

Just as corporations and other organizations use video to communicate with both internal and external audiences, government facilities also rely on video to disseminate information. With governments, much of the video is meant to fulfill access-to-information mandates, such as when a state or local agency captures and records meetings for the public to stream live or on demand. That’s just one example. There are about as many video use cases as there are government agencies. The point is, whether it’s for public access and engagement, training, distance learning, collaboration, emergency management, or promotion, video production and delivery have become a big part of government operations today.

In the 90s, when digital video was in its infancy, governments used purpose-built, hardware-based media storage systems when creating and delivering video. That’s because back then, video bandwidth needs far outstripped typical IT storage systems. In the early days, video operations even required modified hard drives that wrote and read video using all the heads in parallel. Over time the industry moved to RAID arrays that could accommodate first one video stream, then several, and eventually hundreds of compressed streams. Those RAID arrays were sold by media-specific vendors along with specialized software applications (think Apple, Avid, and Autodesk). Quite a bit of testing went into making sure the storage, servers, and software worked together to provide smooth playback and, as importantly, kept up with real-time ingest streams. Fortunately, relying on the alchemy between small media-storage vendors and expensive media storage is a thing of the past. Or at least it should be.

Over the past 10 years, data-access (read/write) rates have evolved much faster than the more talked-about increase in video resolutions, frame rates, and bit rates. Today general IT storage systems can handle the video production and distribution workload for dozens of streams and multiple workstations simultaneously, so there is no longer an advantage to buying media-specific storage. In fact, after decades of focusing on bandwidth and capacity, we’ve come to a point where those qualities, while still important, are no longer paramount for most video operations. It’s easy to get ample bandwidth and capacity from almost any storage system on the market today. Now most operations have shifted the focus to reliability and cost of ownership instead — two areas where enterprise storage systems have a tremendous advantage.

Enterprise storage was designed for organizations such as banks and hospitals that can’t afford to lose their data or have it go offline for any period of time. Enterprise storage manufacturers have spent the past 25 years honing the systems for reliability, uptime, and nondisruptive operations. They’ve done more than simply build more robust hard-drive arrays. They’ve built data-management software on top of those arrays, software that manages for every possible contingency and ensures maximum resiliency. Some enterprise storage systems even offer software with predictive analytics and “phone home” capabilities, whereby the system sends a message to the manufacturer indicating that a drive is likely to fail in the near future. This capability gives manufacturers enough warning so that they can replace the drive before it fails without causing any disruption to the video operation. Because drives will fail. The question is, what happens after they do? The value of having enterprise resiliency built into the storage system’s software layer cannot be overstated. It is the difference between a successful operation and one that is hamstrung by downtime and the inability to grow easily.

For added resiliency, some enterprise systems have also moved from RAID storage, where it can take several days to rebuild data after a drive fails, to dynamic disk pools, where data can be rebuilt almost an order of magnitude faster. Dynamic disk pools also make it possible to replace failed drives during regularly scheduled maintenance instead of having to do it immediately.

In terms of total cost of ownership, there is a misconception that enterprise storage systems are more expensive than media-specific systems. The truth is that from a cost-per-bandwidth and cost-per-capacity perspective, enterprise storage is generally less expensive over the life of the infrastructure than a similarly featured system from a media manufacturer. The key is for government users to take all factors into account when considering a storage system for video. If buyers focus only on initial bandwidth and capacity, they’ll always be able to find a cheaper up-front price for a system, but beware: You get what you pay for. There are hidden costs in maintenance and downtime that not only cause painful disruptions but drive up the total cost of ownership.

Another important consideration for government video operations is the level of security they get from their storage systems. All government agencies, from the military to the IRS to the local DMV, must guard their data closely. Enterprise storage offers military-grade security with almost endless encryption options, including schemes for encrypting data as it’s moving from one system to another and schemes for encrypting data while it’s sitting in storage (data encryption at rest).

For all of these reasons, government facilities would do well to consider enterprise storage for video production and delivery. When doing so, ask these questions:

  • In the event of drive failure, how does the system rebuild the data?
  • Is the data backed up in the system, either through RAID resiliency or snapshots made on site or at a remote site?
  • Can the system essentially heal itself and continue to provide access to the data in the process?
  • Will users have access to the data 24/7 for years, even as technology evolves and media types go from spinning hard drives to solid state drives?
  • Can storage be added as needs grow, without excessive installation or professional services costs?
  • Can the system be managed without relying heavily on a storage technician or storage engineer?
  • Must drives be replaced immediately upon failure, or can replacement wait for a scheduled maintenance window?
  • Are there provisions for proactive maintenance versus reactive maintenance, such as a phone-home function?
  • How and at what points is the data encrypted?

If government buyers ask the right questions during their research — ones that address not just bandwidth and capacity, but all the factors that are important to the operation — they’ll end up with a fast, reliable, secure, and infinitely accessible storage system primed for easy expansion as their needs change.