Sizing It Up
While the move to HD is good from a video quality standpoint, it poses a number of challenges when it comes to recording and storage. Here’s why.
In the world of video recording, products and technologies march forward at an ever-dizzying pace. Many of these changes are being driven by a move to high-definition and megapixel cameras, which in turn is being driven by consumers’ expectation of high-quality video.
If this sounds like the same old song that’s being sung in other areas of the security industry, you’re right. So feel free to sing along at any time.
HD is everywhere: our living rooms, computers, video cameras, even mobile phones. Standard definition? A thing of the past, at least in the consumer space. Security isn’t there yet, but it’s headed that way, says Vince Scarpelli, product strategy manager for video platforms at Boca Raton, Fla.-based ADT.
“HD is becoming the new standard definition. Customers are pretty much asking for high definition and moving away from the standard definition VGA type devices,” he says.
Any discussion of video recording eventually must include a discussion of storage. So while the move to HD is good from a video quality standpoint, it also poses a number of challenges when it comes to storage. Obviously, as resolution increases, so too do file sizes. These larger files necessitate greater storage capacity. And there’s the rub.
But wait, there’s more. As you consider how best to store the larger file sizes generated by HD and megapixel cameras, you also have to consider how those large video files will move not only from the camera to the storage solution, but from either camera or storage solution (or both) to operators’ or end users’ screens. That’s when the discussion grows to include bandwidth.
“As cameras increase in resolution, file sizes have grown exponentially,” says Jeff Whitney, vice president of marketing for Intransa, located in Cupertino, Calif. “This growth has impacted the amount of storage media required in addition to the network bandwidth used to carry the video from the cameras to the recording platform.”
Here’s how it works, as Whitney explains: Video from a single IP camera streams to the storage device as a large, sequential block of video. As more cameras are added, streaming changes into large, non-sequential blocks of video, which can be a challenge for storage products to keep up with. Why is that? Because when the IP protocol is used in the IT applications for which it was designed, it can consume all of a network’s available bandwidth, and will keep re-submitting data until the storage device acknowledges that it has received all of the data.
“Unfortunately, with video, the bandwidth consumption of multiple cameras is such that video frames can be dropped with the IP protocol, resulting in frame loss,” Whiney says.
As we all know, in critical security applications, any sort of data loss — whether one frame or 100 — is simply unacceptable.
Additionally, he says, the traditional record-erase-record cycle common in video systems since the days of videotape can fragment the disks on which video is stored. Over time, both recording and playback performance will decline. Running a defragmentation process will fix the problem, but depending on the level of fragmentation, that can be time consuming. It also takes the drive offline, making any data stored on it unavailable, which isn’t possible in a 24/7 video surveillance and recording situation.
Despite these obstacles, however, there’s a video recording solution out there that’s just right for almost any imaginable application. The trick is to find it. And to find it, you have to know where to look.
Managing the Demand
Yes, compression and other recording algorithms like H.264 have their place. However, Whitney points out, maintaining high quality in both streamed and recorded video has become increasingly important in a growing number of applications and verticals. These include retail, manufacturing, law enforcement, gaming and transportation.
Naturally, as is the case with every installation, the first step is to determine exactly what it is that a customer needs from their video solution and to address issues around high-quality video streams before they’re recorded. One simple question will almost always rule out real-time streaming of high-quality video around-the-clock. Anyone who says otherwise simply can’t deliver on that promise, says Bill Lawrence, director of service solution for ADT’s product management group.
“Clearly, if you’re trying to stream HD-quality video at high frame rates across a typical internet connection, it isn’t going to work. You just can’t do that,” he claims.
What most people find, Lawrence says, is that lower resolution and frame rates are perfect for their day-to-day operations. Higher-quality video is only required in certain situations or under certain conditions. For those customers, analytics can play a major role in not only preserving their bandwidth, but providing them with the highest-quality video they need when they need it. For example, if a perimeter is breached or motion is detected in a certain area, then high-resolution, high-frame rate video will stream to operators’ screens — and eat up precious bandwidth — for only a short time.
“You can use analytics to filter out the things that you don’t want and only capture the things that you do want,” Lawrence says. “Then you can make much better use of the available bandwidth and storage capacity, and I think cloud computing becomes a little more of a reality.”
In fact, the recent maturation of video analytics in general has been a major factor in the drive towards HD and should continue to be so into the future, Scarpelli says.
“Analytics has been stalled for about five or six years, really showing no movement as the marketing hype far outreached the capabilities of the actual analytics themselves,” he says. “Recently, in the past six months or so, we’ve seen some mature analytics come to market — analytics that work fairly well.”
More and more, those analytics most likely will be pushed towards the edge, which should go a long way in alleviating bandwidth demands, Scarpelli says. Between network attached storage and the continued drop in the price of memory, storage should increasingly move to the edge as well.
“As the trend toward HD continues to move forward, this growing trend of edge devices will grow, as well “There’s a big push, and we’re seeing more and more things being moved to that edge,” Scarpelli says. “SD cards, storage, analytics running at the edge — all those things are being pushed off the server and out to that edge as much as possible.”
Other technologies and solutions that will help ease the pain of storing HD video include larger capacity disk drives (which will populate unused hard drive bays) and DVR expanders to supplement overextended hard drives, says Ed Strong, director of the Western Digital CE Products Group in Irvine, Calif.
And then there’s “The Cloud,” which not only contributes to bandwidth and storage issues, but also introduces a whole new set of challenges.
Thanks to Microsoft’s slick advertising campaign, we’ve all heard of “The Cloud.” Some may not understand exactly what that means, but the general idea is pretty clear. And while the cloud may seem like a brand new concept, the only thing new about it is widespread awareness of its potential.
“Cloud computing is something that’s been entrenched in the IT industry for a long time and is starting to move into physical security,” Lawrence says. “There are some obvious challenges with it from bandwidth perspective. For example, how do you get the amount and quality of video that are required and that people are expecting into the cloud in a cost-effective way?”
While adoption of cloud computing hasn’t spread broadly across the industry, there have been many implementations of the technology, and the future should bring even more. The ability to broadcast two or more streams of video simultaneously is going to be one of the main drivers towards the cloud, Whitney says.
This type of simulcasting allows recording capacity at a location to be limited to smaller amounts, such as 2TB or 4TB. These act as a primary recording buffer, while additional streams of video can be transmitted to multiple locations for review or consolidation. Often, Whitney says, that off-site location is an IT-grade data center where server farms store video data from many sources in a single location, allowing storage capacity to be managed much more efficiently than with multiple storage platforms.
“This type of early implementation of cloud technology is becoming more common,” Whitney says. “The key to its success will be having reliable, modular appliance platforms at the remote locations with matching devices capable of decoding the video for use on standard IT storage platforms to eliminate fragmentation and frame loss.”
At the same time that awareness of the potential of cloud computing is growing, the cost of bandwidth is shrinking. These two factors should converge in the near future to enable cloud computing for video applications to become more reliable, Lawrence predicts.
Perhaps the greatest barrier that is inhibiting greater adoption and acceptance of cloud computing is, oddly enough, security. Logical security, that is. Many wonder just how secure data stored in the cloud really is. Most of that, Lawrence says, is perception. (See related article, “The Cloud: Secure Enough for Security,” on page 107 in this issue.) When Sony’s gaming center was recently hacked, thousands of customers’ personal information was compromised, which was big news for a while. So even if perception is the major obstacle, it’s not enough to brush those concerns aside.
“It is a perception, but you can’t just ignore that because it is a very real concern,” Lawrence says. “And particularly when you get highly publicized breaches like Sony, you’ve got to deal with it. You can’t walk in the door the next day and say, ‘Here’s a cloud service for your mission-critical video.’ That’s going to be a hard sell.”
While breaches like Sony’s help create a greater fear of data security in the cloud, it’s worth noting that customer-operated systems are compromised all the time. Even the U.S. Government, which employs the most state-of-the-art security measures, has its own problems with being hacked. So there’s nothing that’s truly 100 percent impenetrable, but there are some very solid information security standards that are constantly evolving to address new concerns, Lawrence says.
“There’s a notion that if it’s in the cloud, it’s more accessible, but in some ways, everything is in the cloud because it’s all touched by the internet,” Lawrence says. “If you think about it, your data in the cloud is no more vulnerable than it would be if you stored it on your own personal site. In fact, it may be even less so because we’re using commercial-grade security, and the average business probably is somewhat less vigilant.”
So could The Cloud reduce storage concerns to minor details? Only time will tell.
It Really IS What You Know
In today’s market, installers and integrators have to have some network savvy to remain competitive. This is a truth that’s becoming more universal across more segments of the security industry — including storage.
“When you talk about storage, you’re really talking about a whole computer. You’re talking about operating systems and you’re also talking about video management software in a lot of cases,” Scarpelli says. “So you’ve got to understand all that terminology to start with, plus understand how all those things come together. And it often is not just a simple ‘plug it in, turn on the switch and you’re done.’”
Western Digital’s Strong says the growth of high-resolution video technologies will lead installers and integrators to give more thought to how many drives a job will require and what capacity those drives should provide. They should also consider specific types of drives in selecting storage for a video system.
“When typical, non-surveillance-grade drives are installed, they fail since they’re designed to work for much shorter periods under less-demanding circumstances,” he says. “The overall costs associated with operational downtime, replacing failing hard drives and dealing with any data loss and recovery issues can far outweigh the price difference in opting for surveillance-grade storage from the outset.”
For its part, ADT has made sure that its installers, integrators and other key personnel are getting the necessary training and certifications (Cisco, Microsoft, et al) they’ll need to succeed with storage and other IT-related technologies that have become increasingly popular in the industry. Scarpelli’s advice for other installation companies is not only to provide that training, but to look at employees’ knowledge and skill sets much earlier.
“Perhaps companies’ hiring practices need to change, so that they’re looking at technicians who come out of computer courses such as Cisco’s and Microsoft’s as opposed to, say, electrical and technology courses,” he comments.
|How Much Storage Is Enough Storage?|
A general rule of thumb for determining the necessary amount of storage would be helpful, but it doesn’t exist. The large number of variables involved in video surveillance makes a blanket guide impossible, Scarpelli says.
“You really have to depend on the different manufacturers’ calculators to figure out what that storage is,” Scarpelli says. “I wish there was one overall rule of thumb, but we really don’t have one.”
That’s not to say finding the right storage is a shot in the dark. For short-term retention, 2TB is a good starting point, Whitney says. Beyond that, it’s important to ensure that a solution is scalable to allow for additional recording capacity — at the same speed, without halting system operation or requiring any rewiring.
Calculating actual recording needs, Whitney says, is all about accurately capturing the system requirement. This involves factors like the number of cameras, the frame rate and resolution of each, percentage of motion, hours of operation and desired recording period. Entering this data into a manufacturer’s calculation tool makes the actual calculation extremely simple.