Imagine you are on a trip and decide to record a video on your smartphone or tablet. About two minutes into recording, you receive a push notification that you have run out of storage. Now, multiply the length of that video by 720 times to get a piece of footage 24-hours long — recorded not just on one device but by hundreds or even thousands of cameras at once. What you have is the greatest challenge faced by video surveillance organizations today: where to store all that video footage.
Video is the fastest growing data type, used for everything from monitoring retail traffic patterns to conducting crime scene investigations. As organizations continue to change how they collect data, their storage needs evolve as well. Storage has become an expensive, moving target. In fact, when embarking on a video surveillance project, an organization can expect to commit up to two-thirds of its budget to cost of storage.
New laws and regulations, cameras with higher resolutions, and new VMS packages with analytics capabilities are important developments, but each creates new data that must be stored. As the volume of video data grows, so does the challenge to store it. The time has come to address this as an industry, to look to new technologies, and to evolve as media and entertainment did to serve our organizations and our customers more efficiently than ever. With these challenges in mind, here are three video surveillance storage trends to watch:
New Laws and Requirements
In addition to a general increase in capacity needs, there are two types of new laws and regulations leading to the expansion of video surveillance data in the market.
The first type of law mandates the creation of video surveillance data. A good example is a recent Texas law that requires schools and special education centers to record video throughout their facilities. These laws create a substantial amount of data and a new storage need that did not previously exist.
The second type of law mandates retention of video surveillance data for a length of time. The Cruise Vessel Security and Safety Act (CVSSA) of 2010, for example, requires the cruise line industry keep video data for a minimum of 120 days. Financial institutions and ATMs have a 180-day retention requirement, along with retail and grocery stores, which must also retain their surveillance video due to potential legal obligations. Police departments must retain surveillance videos from interview and interrogation rooms for a duration longer than 999 days. In military detention centers, videos are retained for up to seven years.
These requirements protect organizations in the event of a lawsuit, allowing them to build a case with video surveillance data on file, and to protect the public in case of incident. However, they also dramatically increase an organization’s need for and cost of storage. To ensure new laws and retention requirements help instead of hurt, integrators and dealers need the ability to offer affordable storage solutions designed specifically for video.
New Technologies for Video Storage
There are two kinds of data in a traditional IT space: structured and unstructured. Structured data, like a database, has specific formats and patterns. Current IT storage technologies are designed for structured data management and do it well. An example is disk with deduplication (“dedupe”), which finds patterns in numbers and letters, and reduces redundancies by cataloging duplicates in metadata.
Video is unstructured data, made up of innumerable, unique pixels with no format or pattern; and thus there is no way to use metadata to reduce the size of video data sets and storage solutions that work well in common IT environments do not translate well to video surveillance.
With video surveillance, scalability and cost are persistent issues. To compensate, organizations are reducing the amount of video data created by skimping where they can — they buy 1080P cameras, but run them at 720P. They want full motion with 30 frames per second but settle for 5 to 10. Percentage of motion suffers as organizations shoot for 50 percent and end up doing less. The result is low quality, grainy video, frame skipping and lag.
Back-end storage technology drives these decisions, rather than quality, safety, or even ROI. This is clearly a case of “tail wagging the dog.” The question should not be “what is your video retention requirement,” but instead, “what would you like your retention policy to be?”
To address this, we need a different kind of storage technology for video surveillance, and this evolution has happened at least once in the past. The industry is going through the same technology shift that the media and entertainment business did a decade ago, when media and entertainment looked at shelves of analog film reels and decided it needed two forms of storage to manage the practicality and cost of storing unstructured data.
Today, the industry keeps a small percentage of data on spinning disk for round-the-clock access and archives data to digital tape as a more affordable, dense form of storage.
Diversifying storage mediums with a central access point is the ideal infrastructure for video surveillance, enabling customization to an organization’s unique needs, requirements and frequency of access. A starting point might be Shingled Magnetic Recording (SMR) disk, a technology specifically designed to handle large file types and sequential data. SMR disk is highly scalable in both capacity and performance to accommodate additional cameras and retention requirements, and provides long term IP video data storage at the cost of tape.
SMR coupled with a tape system can enable customers with large video retention requirements to keep costs down while maintaining a tiered infrastructure for data preservation.
The Rise and Rules of Cloud for Video
As in other storage verticals, discussion of public vs. private vs. hybrid cloud is a hot topic in video surveillance. Security is usually the first concern; however, the threat is generally equivalent across all three types of cloud, with public cloud being perhaps more vulnerable due to offsite infrastructure.
When looking at the cloud for video surveillance, there are three main things to consider:
1. Chain of custody. Many organizations generating video surveillance data, including police departments and governmental agencies, have chain of custody requirements. Since it is difficult to prove chain of custody when video data has been touched by someone outside the organization, cloud storage can be tricky.
2. Cost. When calculating the cost of cloud storage, organizations must include the cost of withdrawing data if a cloud vendor closes its doors, or if the user decides to change cloud providers. Video data in particular — because it is unstructured — faces a high cost of transmission. For these reasons, full cloud solutions are generally cost prohibitive and non-viable for the video surveillance market.
3. Proprietary nature. Proprietary nature describes the extent to which organizations have sole ownership of and fast access to their data. Ownership enables organizations to access data at a moment’s notice, rather than waiting for a cloud provider to withdraw the data for them.
For genetic diversity in the data center, a hybrid cloud storage model is recommended. Hybrid cloud uses a mixture of on-premises, private and public cloud storage to create a seamless and secure solution, customized to each organization. A combination of private cloud with the industry-accepted SHA-256 hashing algorithm and on-premises disk storage with dedicated hardware is the best of both worlds, ensuring security, reduced susceptibility to data breach, chain of custody and immediate access.
The use of multiple storage technologies is also vital to data preservation in the event of cyber attack, outage or natural disaster.
Brian Grainger is the chief sales officer of Spectra Logic. Follow him on Twitter, @BrianonStorage. Request more info about Spectra Logic at www.securityinfowatch.com/10797249.
