Video storage has unique requirements

No "one-size-fits-all" device meets the requirements of every video application.

By Rich Bourdeau

—Generating video content is no longer the exclusive domain of television and movie companies. The Internet is exploding with video and other rich media content. Security and regulatory requirements are causing a rapid expansion in the use of video surveillance. Imaging, video, and other multi-media content is quickly making its way into increasingly more mainstream business applications across all industries. The rapid growth of video applications is driving the need for inexpensive, simple to use, highly available, and scalable storage. As these video applications proliferate, they are becoming larger and more mission-critical. Current data storage infrastructures used for these applications no longer meet their needs.

This article discusses the unique requirements of these applications, looks at some alternative deployment models, and highlights why clusters of IP-connected storage based on industry-standard components are rapidly becoming the storage deployment platform of choice for companies looking to upgrade or deploy new video applications.

If pictures are worth a thousand words, then videos allow people to communicate with each other in even more-effective ways.

As a consumer, the options for entertainment and education with video programming are almost endless. What used to be a three-channel world has quickly evolved to hundreds of specialty channels with on-demand capabilities that allow viewers to watch whatever and whenever they want to watch. To feed this rapid growth, the broadcast industry has moved quickly from analog and tape deployments to digital infrastructures.

Just the growth in broadcast video and movies alone would generate a healthy demand for more storage capacity. However, with the additional growth of digital video content being distributed and shared across the Internet, the amount of video storage will soon eclipse more structured data, if it has not already done so. User-generated content is evolving from text and pictures to music, videos, and other rich media content. Made possible by the availability of inexpensive network bandwidth, this new video communication medium consumes far more storage than more traditional applications.

Videos are not limited to just entertainment and the Internet. They are also finding their way into more traditional business applications. Security, time and motion studies, traffic and transportation, consumer behavior analysis, and supply chain monitoring are just a few of the applications of video monitoring technology. Storage for video surveillance used to be something that was recycled every few weeks. Now, with increased regulations, liability protection, additional uses, and post-processing analysis, video surveillance data is being archived for longer periods of time, at greater resolution, and from more sources than ever before.

Spurred on by the rapid expansion in network bandwidth and increasing storage densities, multi-media content is increasingly becoming part of more and more business applications. In a sound-byte society, where people want their information quickly, there will be no shortage of applications driving the capture and sharing of more digital video content than we can fathom.

Over multiple decades, the largest Fortune 1000 companies have grown their storage infrastructures to hundreds of terabytes, with the largest companies having multiple petabytes. In less than a year, Web 2.0 Internet start-ups and other companies storing rich media content have consumed more storage than the larger companies have accumulated over their lifetimes.

Social networking, application hosting, auctions, photo sharing, and video distribution all consume more storage than older transactional type applications. The difference is that many Web 2.0 applications tend to be participatory and collaborative applications, where the users are the ones generating much of the content. The change from provider-generated content to user-generated content makes predicting storage growth far more challenging than in the past.

Since the Internet never closes, these new applications require many of the same high-availability and data-protection features of their mission-critical predecessors, but at a price point and management simplicity not possible with the more traditional offerings of the leading storage providers.

In addition to consuming vast amounts of storage, these applications often have different I/O profiles compared to transactional applications, which tend to be primarily small- block, random I/O-oriented. Web 2.0 applications tend to store larger, more variable digital content. Access patterns can also vary between read-intensive (video distribution) to almost exclusively write-intensive (remotely hosted backup) and anything in between. The storage architectures that supported high-performance transactional applications are not always a good fit for new video applications.

As for scalability, it's not that traditional monolithic arrays can't scale up to petabyte-plus configurations; it's the efficiency and cost (capital and operational) required to reach these levels. Not everything scales gracefully when 10-to-20-year-old storage architectures strain to accommodate the scalability requirements of these new applications. Management paradigms that worked fine with tens of terabytes don't always scale well when managing multiple petabytes or more.

Like storage in general, there is no "one-size-fits-all" device that meets the requirements of every video application. However, video applications do share some common storage requirements.

Budgetary constraints
Due to the large amounts of capacity consumed by video applications, the cost per terabyte needs to be far lower than price points of traditional SAN and NAS storage devices. While companies would like cheaper storage for all of their applications, many video applications would not be financially feasible without inexpensive storage. For most video applications, the price point required to make their applications viable is in the range of $1,000 to 2,000 per TB with prices declining at 30% to 40% per year.

Even within the video storage market, inexpensive storage is still relative. Applications such as broadcast video editing and on-demand video distribution may be able to afford more expensive storage than social networking and video-surveillance applications.

Video applications have different requirements and profiles than traditional data-center applications. Traditional data-center applications consist of lots of random, small-block reads and writes. However, video applications typically consist of large-block, sequential streaming I/O, with unique I/O profiles compared to traditional IT applications. For example, many video capture applications are 99% writes, while video sharing applications are 99% reads.

Over the years, traditional storage has been optimized primarily to meet the needs of transaction processing applications. The use of cache was a big part of achieving optimal performance. Large block, sequential reads or writes can quickly overwhelm cache, rendering traditional storage systems much less effective for use as video storage.

Video applications, on the other hand, benefit from storage architectures that look to distribute video data across multiple disks and servers to increase the degree of parallelism. This approach not only provides a more cost-effective implementation, but it also allows video-optimized storage to meet the performance and throughput requirements that might not be met with more traditional storage architectures.

Operational simplicity
Historically, video storage came as part of a media solution. Embedded with an application such as video editing or video surveillance, these applications often live outside the domain of IT management. The people administering these applications don't have the expertise of traditional IT staff. Therefore, management of the storage has to be far simpler than traditional storage. However, with the increase in the use of video applications of all types, companies are starting to consolidate management of video storage within their IT departments.

As video storage shifts under the IT management domain, due to the sheer volume of data, companies cannot afford the operational overhead of complex management paradigms.

Just as acquisition costs can jeopardize the viability of a video application, so can excessive operational management costs. Looking at a four-year total cost of ownership of a video storage asset, the cost to manage a TB of storage is often overlooked, but can often be as high as four to five times the initial acquisition costs. Regardless if video storage is being managed by a line of business or IT department, it needs to be as simple as possible to manage. New storage technologies, which simplify and automate the configuration and ongoing operation, are critical to video applications, which consume from hundreds of terabytes up to multiple petabytes of storage.

Modular growth
Video applications consume large amounts of storage. However, the pace of growth is not always predictable. This is especially true of applications with user-generated content, but even more traditional video applications can experience unexpected growth due to expansion in the number of data sources, higher resolutions, or storage in different formats to facilitate distribution through different mediums.

One approach is to purchase a large monolithic storage array up-front, which will accommodate both planned and unplanned growth. The problem with this strategy is the monolithic arrays typically cost significantly more than the identical capacity in a modular array. In addition, the acquisition cost needs to be allocated up-front, when capital is usually limited and the ROI or viability of the business is still unproven.

Modular storage provides a pay-as-you-grow storage model. However, modular arrays, with clustered or grid architectures, have several advantages as a platform for video storage when compared with traditional dual-controller modular arrays. With traditional dual-controller arrays, additional capacity can be added up to a point. However, when bandwidth or controller processing power runs out on one of these systems, then a fork-lift upgrade must be performed or another modular array added, with each additional array being managed as a separate entity.

Modular storage arrays with clustered architectures allow capacity, bandwidth, and processing power to be scaled in modular units. Storage virtualization software eliminates the operational complexity of having to manage multiple independent arrays by creating a single virtual array that leverages the combined resources of the cluster to provide more linear performance and scalability. The array's virtualization software takes care of automatically distributing the data and processing across new modules as they are added. Clustered storage architectures provide pay-as-you-grow scalability while maintaining the management simplicity of managing a single entity.

Most video applications require fault tolerance to allow for failure or upgrade of individual storage components without disruption to the application. In addition to redundancy of hardware components, RAID-type data protection is also a requirement. However, many video storage implementations are moving away from traditional RAID protection to other protection schemes that provide faster rebuild times in the event of a disk failure or replacement. As capacities grow, the chance of a disk failure increases with the number of disks. The ability to rebuild quickly, without impacting performance over long periods of time, is an absolute must.

Disaster recovery is typically a function of the importance of the application to the business. While it is important to keep all video applications running, some applications—such as video monitoring—may be able to tolerate downtimes of hours or days, whereas broadcast video or video-on-demand distribution may require the same resiliency and availability as a financial application. Disaster fault tolerance is one feature that differentiates the offerings and cost of video storage vendors.

Video intelligence
Intelligent services are built into most storage arrays. Storage services, like volume management, snapshot, point-in-time copy, and remote replication, are services that companies have come to expect in their storage arrays. Most storage used by video applications has some or all of these capabilities, depending on the target applications.

In addition, some arrays have features specific to video applications or features that optimize video storage and retrieval. For example, both Blackwave and Omneon can create multiple copies of frequently accessed videos that allow for more-efficient content distribution. Compellent and Blackwave automatically move frequently-accessed data blocks and files to the outside of disks to improve disk seek times and performance. Another example is the automatic format transcoding available in Omneon's storage systems. This type of storage allows video content to be converted to different video formats, automatically creating new revenue possibilities by using the same video through different distribution mechanisms.

Video ecosystem
While almost any storage array can be used to store video, some vendors have optimized their systems for specific video markets. Their hardware- and software-compatibility matrices look quite different than most storage vendors.

For example, Intransa and Pivot3 have focused on the video monitoring and surveillance market. They integrate and support camera, management systems, and video analytics that target the specific requirements of this market. Avid and Omneon are examples of vendors targeting the video editing and broadcast market. In markets like these, storage is often embedded with applications, making it easier for most customers to consume.

Direct-attached storage (DAS)
Due to a combination of large amounts of data and low initial acquisition costs, many video applications still store their data on inexpensive DAS. This implementation approach works well for companies with small amounts of data, but for most video applications, DAS-based storage solutions quickly become multiple islands of storage that are hard to scale and complex to manage.

DAS represents one of the most basic storage deployment options, where disk drives are either embedded in or directly attached to a single server. To access storage, other computer devices must be networked to the server with the storage.

DAS storage models still comprise a large percentage of the video storage market, especially in price-sensitive applications like video storage.

Increasingly though, many companies are running into the scalability, availability, and performance limitations of DAS storage. The initial cost savings of DAS can quickly be consumed by the operational costs or business impact that result from these DAS limitations. As networked storage addresses the specific needs and cost requirements of video applications, the Taneja Group believes that IP-connected storage will emerge as the dominant deployment model for all types of video applications

Traditional SAN and NAS
Traditional Fibre Channel SAN or NAS storage devices can be used to store video data. If a company has a relatively small amount of video content and a large investment in existing SAN or NAS technology, then these storage devices may provide a cost-effective approach for video storage without having to retrain operational staff to manage new storage architectures from different vendors.

However, if the objective is to store large amounts of media-rich content with high-throughput requirements, then traditional SAN and NAS storage devices will not be cost-competitive compared to newer clustered storage architectures. In addition, traditional storage devices will likely run into performance, scalability, and manageability limitations that will further exacerbate the cost disparity.

Clustered IP storage
IP-based clustered storage represents a cost-effective alternative that meets the unique requirements of video applications. Based on commodity servers and disk drives, clustered storage systems have many of the cost advantages of DAS. However, clustered storage can aggregate multiple storage nodes into a single logical system. Using this modular approach, performance, capacity and bandwidth can be scaled incrementally as needed while preserving the simplicity and manageability of managing a single virtual storage device.

IP-connected storage provides the bandwidth and performance required by media-rich applications without the additional cost overhead and operational complexity of deploying Fibre Channel SANs. Storage applications are using iSCSI (block-level), NAS (file-level), and object-level interfaces for the storage and retrieval of video content. Each of these interfaces has its advantages and disadvantages, and depending on the application, one may be more appropriate than another.

Overall, clustered storage architectures are well-suited to meet the demanding requirements of video applications. Even within video storage, there is no "one-size-fits-all" solution. Most clustered IP storage solutions can meet the needs of generic video applications. In addition, some vendors have also chosen to address the needs of specific markets such as video surveillance (Intransa and Pivot3), broadcast video (Avid, Isilon, Omneon) and video distribution (Blackwave).

Storage as a Service
One option for companies looking to store video or other multi-media data is the Storage-as-a-Service (SaaS) delivery model. Amazon S3 and Nirvanix SDN are examples of companies offering a Web services interfaces for storing media-rich content for all types of Web 2.0 applications.

Vendors that deliver SaaS storage services provide a simple Web-service-based API to store and retrieve multi-media objects in their hosted storage. Costs are in the $1,500 to $2,500/TB per-year range to store data, with additional costs each time data is transferred in or out. While these rates may initially sound expensive, when factoring in the fully loaded costs to house, cool, and manage in-house storage, SaaS storage services may be a financially viable alternative depending on the data access requirements.

Many Internet start-ups prefer this pay-as-you-grow model of SaaS storage services. They would rather invest their limited capital and technical resources in development of their applications, rather than deploy their own storage. Therefore, Internet start-ups or companies looking to pilot new video applications are more appropriate consumers of SaaS storage services than traditional large enterprise customers.

Taneja Group opinion
In the coming years, video and other media-rich applications will continue to consume a larger percentage of every company's storage resources. Most companies, whose primary business involves the creation or distribution of video content, have already realized that traditional storage architectures did not address the cost, performance, availability, and manageability requirements of their business. Faced with limited options, many companies started building their own video servers from commodity components or purchased high-end media servers.

For companies whose primary business is not video applications, traditionally media data was kept on low-cost storage that came embedded in video solutions. IT was not typically involved in the acquisition or ongoing management, because this information was not deemed "strategic." However, as the amount of video data being captured as well as its strategic importance grew, many companies started looking for more cost-effective alternatives for storing this rapidly expanding digital asset.

The Taneja Group believes IP-based clustered storage provides an attractive alternative for the storage of video and other media-rich content. Due to the sheer volume of video data being generated, traditional DAS, SAN, and NAS solutions are just not economically viable. We believe that IP-based clustered storage is poised to garner the lion's share of the video storage market. However, once customers become accustomed to the technology and realize its cost advantages, we would not be surprised if it not only captures the video storage market, but also makes inroads into storage used for more traditional applications.

If video is your game, evaluating some of these newer storage vendors will provide you with attractive alternatives to JBOD storage at the low-end and traditional dual-processor controllers or monolithic storage arrays at the high-end.

Rich Bourdeau is a senior analyst with the Taneja Group research and consulting firm.

Clustered IP storage vendors:
3PAR, Blackwave, Compellent, Dell EqualLogic, Intransa, Isilon Systems, Omneon, Pillar Data Systems, Pivot 3, XIV

This article was originally published on May 24, 2008