Download the authoritative guide: Cloud Computing 2018: Using the Cloud to Transform Your Business
We surround ourselves with data that we need to store. Think of your “To File” manila folder or the sea of papers you navigate to get to your desk every day. But also think of the vast amount of digital data you need to keep track of. You generate some of it yourself, while others send you a whole swamp full of information to read, use, edit, store, study, cite, and so on, both in-house and on the Web. Now multiply that by every person in your enterprise. Wow!
While our present-day needs for data storage are pretty dramatic, we’re guessing the future is going to be even bigger and better — and crammed with even more data to store. We recently had the opportunity to talk about the future of enterprise data storage with Tom Coughlin, author of the 2004 Entertainment Digital Media Storage Report and the force behind the Storage Visions Conference.
We thought he’d see the far horizon of data storage based on the intent of this conference, held in January of each year, which is to bring together industry visionaries, content providers, storage system providers, and end users to focus on the intersection of data storage with content creation, content delivery, and electronics systems.
Marty Foltyn (MF) and Margaret Akin (MA): Tom, let’s start off with the big question. What is the top challenge in content storage?
Tom Coughlin (TC): First of all, we have a lot of drivers creating digital content that needs to be stored and manipulated. We’re converting existing analog content to digital, plus we’re creating and handling new digital content. The content value chain is a circle that starts with creation, then goes to editing, then archival holding and distribution, and ends with final reception. Important elements that drive demand for storage and capability are all increasing.
MF: So, what’s in the circle to help us store the content?
TC: Storage devices offer an interesting value proposition to users. Although the growth in areal densities (megabits per square inch) has slowed, capacity is already enormous and is still increasing. For example, we’re looking forward to serial ATA (SATA) disk capacity of over 1 terabyte on a single disk drive in three years.
MA: What about the architecture of our storage systems?
TC: With more and more human content available online, we also need storage support for transaction processing and delivery. This trend favors advanced performance through Fibre Channel, serial SCSI, SAN, and NAS. To keep the content available, we need to use in-line and near-line storage before we move the data to archive storage. Most archiving is done for preservation driven by regulatory reasons as well as the preservation of entertainment content. In general, this is the province of tape and optical media.
Continued from Page 1
MA: Which type of storage will we want the most of?
TC: The highest growth will be in lower-cost storage. While it’s true that editing systems, for example, use higher performance storage, after the content is created, the content is relatively static, so lower cost ATA is an excellent option.
Let’s look at one slice of data storage: archiving. Just archiving all the content is a humongous job. Tape offers the best value proposition at this time for archiving, as technology development lets us increase the capacity available per cartridge. However, because of the sheer volume of the need, other storage technologies may tap into the archiving market.
MA: What about delivery?
TC: Systems need higher performance for the initial delivery of content. If we use a network for content delivery, we can combine storage technologies. While doing video downloads requires a high delivery rate and favors transaction processing systems, the content behind the interface is going to ATA systems. It’s cheaper. We just need a big buffer for downloads.
MA: How will mobile fit in?
TC: The enterprise will grow to combine static content storage systems with the requirement of a large mobile workforce. Richer content for enterprise mobile workers will require larger local digital storage. One driver may be mobile phones with some sort of digital imaging capability. In the U.S., phones with digital still cameras are just becoming popular. In Japan, digital still cameras built into phones have been very popular, and we are now seeing phones on the market in Japan with video camera capability.
Available content on the front end increases demand for that content and the required digital storage on the back end. Remember the content value chain that shows the circle connecting those who use content to those who make content in an ever-growing cycle.
MA: Will phones get bigger to handle the video their new cameras are capturing?
TC: Good question. When the corporate customer takes a moving image of something with a phone, the image requires some significant storage. The enterprise is pushing for disk drives inside the phone. Toshiba recently announced a 0.85-inch disk drive for embedded storage, which I believe is targeted at the nascent video camera cell phone market.
In the near future we will gain mobility and access to customers without reverting to brick phones. Small form-factor storage is increasing corporate communications. With a sufficient broadband connection, a couple of video camera mobile phones can act as video phones. This could be the realization of a 60-year-old technology dream.
MF: Will video in the workplace have an impact on storage requirements?
TC: Content creation for HDTV, uncompressed, demands up to a 10-fold increase over standard TV. Users require larger storage for high definition digital video recorders (DVR) and personal video recorders (PVR). These are perfect tools for the enterprise to deliver education, training, video help, video installation, and other products that can benefit from a visual delivery. And as the workplace offers these high-caliber products, the storage infrastructure behind them will need to keep pace. They all feed upon each other. Changes lead to growth, which leads to further changes.