Storage Peer Incite: Notes from Wikibon’s July 03, 2007 Research Meeting
This week Wikibon presents iPhone demands storage solutions. Apple's introduction of the first handheld device with 8 GB of memory, more than most desktop and laptop computers, plus cellular and WiFi connectivity, and a fully functional browser that can access most Internet Web sites directly rather than depending on special services provided by the cellular service provider as many smartphones still do today, is potentially revolutionary. And this may only be the start. The video iPod has a 75 GB hard drive, implying that a smart phone with 4-8 GB of memory and 75 GB of hard drive could be around the corner. While the iPhone's software is in many ways immature, we expect Apple to add significant functionality over time. We also expect other smartphone makers to step up to Apple's hardware challenge and field their own handheld devices with hardware capabilities similar to the iPhone's, capable of supporting more robust operating sytems and user interfaces and a huge leap in handheld capabilities in general, turning handhelds from techie toys into communications and productivity appliances. We do not see these as PC replacements any more than PCs replaced servers or mainframes. Rather they are a new class of device. From the storage standpoint, they will generate huge amounts of data, much of it video and audio, that will need to be captured, stored, backed up, restored when the device fails or is lost, and archived. This edition looks at some of the storage implications of the maturation of smartphones into fully capable tiny computing devices. Dave Vellante
Last week Apple made its seminal announcement of the iPhone. We observed elsewhere that the iPhone is likely to have a significant effect on how users conceive important classes of enterprise applications (e.g., the availability of a browser pushes the need for high quality Web services). We also observed that the administration of those devices will be closely tied to security concerns.
However, we see a second domain where enterprises will have to clearly articulate their IT infrastructure strategies for accommodating handheld computing devices with built-in phones like the iPhone: storage solutions.
In the last few years large segments of the user, business, and supplier triumvirate have recognized an oncoming transition in how storage is conceptualized, purchased, implemented and administered. These groups have had different agendas, which often created confusion over the roadmaps and even the objectives for storage. However, ultimately all parties agreed that sound storage strategies focus on four key issues:
- Gaining greater control over the rate of growth and price of storage,
- Ensuring that storage fully supports business requirements for quality and compliance,
- Supporting solid and executable disaster recovery programs,
- Providing simple and effective backup/restore capabilities.
We see the iPhone as an important catalyst for cutting through the confusion regarding how to pursue these important storage deployment objectives. By its nature, a very powerful mobile device with large amounts of permanent storage like the iPhone fuels the need to gain greater influence if not outright control over how data is moved through the organization from a storage standpoint.
While we don't see any hardware silver bullets (e.g., homogeneous three-tiered storage solutions), we do see the iPhone fueling the need to invest in practices like information management, technologies like data deduplication and storage virtualization, and business relationships to provide such external services as network-ready hosting of storage for caching and storage delivery.
Action Item: The effects of the iPhone and later devices will take years to mature. Near term, however, storage professionals should factor their potential effects into current architectural decisions both to avoid introducing obsolescence into current investments and help sell the need for high-quality storage capabilities to the business.
Hand-held mobile computing devices can go anywhere and increasingly, as evidenced by the iPhone, will be able to do anything. Moreover, hand-held computers are likely to breed like bunnies, ultimately outnumbering PCs by at least three-to-one (according to various market research estimates). Of course, billions of hand-held computers doing almost anything anywhere means millions of computers breaking anywhere at anytime. Thus, while most envision the wonders of being able to view YouTube anywhere and at anytime, storage managers translate the consequences of these new devices differently: billions of computers doing almost anything that must be backed up and restored (B/R) anywhere and at anytime. The consequences of hand-held device B/R will include new types of administration software, adjustments to data aging and archiving practices, better defined data classification rules and processes, and extensive user training, among other things. However, while many of these consequences may not start to take clear shape for 24-30 months, the "anywhere, anytime" fact of hand-held computers will force B/R to learn how to use public infrastructure extensively. Generally, storage administrators have enjoyed the luxury of being able to operate B/R solutions over wholly or mostly private networks, such as direct-channels, corporate backbones, or leased virtual private networks (VPNs). In the past decade, corporate B/R services have been complemented by third-party services for PC B/R, but increasingly PC B/R has been moved in-house. However, the sheer scale and scope of emerging hand-held computing will force storage managers to employ public networks to operate required B/R services, which imposes a new class of security, performance, contracting, and cost requirements on business's ugliest, critical application.
Action Item: The emergence of powerful hand-held computers will force storage administrators to employ fully public network infrastructure (PNI) to operate global backup and restore services. Storage professionals must factor the characteristics of PNI into their storage strategies immediately, or face further encroachment on the hegemony of the storage function by third-parties.
One of the most interesting debates is whether small mobile computers (SMCs) will require PCs and/or docking stations as they evolve. Will the next generation of users, used to writing assignments, illustrating them with video and pictures, and submitting them at 11:59pm from wherever they are at the time, need anything but the net to support them?
The tired eyes of baby boomers and the cohorts that closely follow them will certainly need larger screens and keyboards. Will users take those screens with them, or will they be ubiquitous in hotel rooms and offices?
All models of usage will be tried, and all will be suitable for some users and some organizations. Looking at how kids behave, the net-supported SMC is likely to become a lifestyle statement for many users.
Action item: Designers of applications and storage strategies will need to assume that many models of SMC support will be in place, including PC, laptop and docking station. Net-only support is the most challenging, and supporting this well will bring with it more effective support for PCs and laptops.
PC desktop storage devices are difficult to integrate into company governance, audit and disaster recovery storage processes. Laptops are more difficult. Small mobile computers (SMCs) will be the most difficult, whether the users are customer, partners or employees.
There is likely to be more than one model of support for SMCs as discussed in the alert “Will iPhones need a PC or docking station?”. Whatever the model, users will be highly dependant on data being held by the web application for both security and performance reasons. In addition, users of these devices will assume that “data tone” equals “dial tone” and will be available 24 by 7 with no break of service.
Application and storage strategies will need to move rapidly towards accommodating these requirement, as the iPhone and the devices that follow in its footsteps gain traction. These will need to include the assumption of zero backup, data administration or application maintenance window.
Action item: Storage strategies will need to evolve to accommodate the always-on 24 hour access requirements of a new breeds of mobile computer, and a new breed of user. Designing strategies round SMCs will have the positive effect of improving distributed support for PCs and laptops.
If one were to argue that mobile devices like the iPhone re-shuffle the priorities of storage infrastructure, it would be easy to justify the creation of a group of storage professionals responsible for ensuring that this storage infrastructure is in place to accommodate backup, restore, compliance and other services for these emerging products. However, while such devices are likely to heighten the importance of appropriately federating distributed computing capabilities, including personal computers, they are unlikely to dramatically alter the pursuit of today's fundamental storage objectives.
Gaining control of unstructured data, providing appropriate backup, archiving and restoration capabilities, ensuring compliance and security (including physical security) and supporting business continuance and disaster recovery imperatives, all at lower costs, remain the core focus of storage administration groups. Organizations must make sure that the skills are in place to meet these objectives regardless of the devices that are creating and consuming storage.
Nonetheless, there is little question that devices such as the iPhone are transferring more organizational power to network infrastructure professionals as they increasingly will be in front of the CIO, along with security teams, discussing what do do about the growth of these new computing devices. Storage administration groups should expect increased tension with networking agendas as devices like the iPhone become more commonplace in the enterprise.
Action Item: Storage administration groups must lead, follow or get out of the way when it comes to providing infrastructure to support devices like the iPhone. The most logical way to lead is to establish clear priorities and strategies to meet primary storage goals independent of the access device and deliver storage infrastructure that supports federated and increasingly mobile user populations with no obvious weak links.
In today's 7/3/2007 Wikibon Peer Incite research meeting, the consensus was that iPhone and similar devices will both consume vast amounts of storage and create demand for more enterprise storage services. Locally, such devices are inherently more attuned to creating storage in the form of video and audio and more cumbersome (than alternatives like PC's) when creating unstructured alphanumeric storage.
Nonetheless, as processing and storage become more distributed, the state of the application becomes harder to manage and lots of experimentation will take place over the next several years regarding whether states are managed locally or globally. This will lead to new questions about what should be synchronized and how it should be done so that data is not lost. As users become more mobile and storage more distributed, the degree to which mobile device storage requires backup, restore and archiving servcies will be a key point of discussion in the coming years.
Action Item: Tier 3 storage is a big winner to the extent that devices like iPhone are adopted in the enterprise. Demand will grow for solutions that provide backup and archiving services for mobile device data in a repository that is accessible, recoverable, low cost and secure. Opportunistic vendors will sieze this opportunity and provide solutions that capitalize on this trend.