"Communities of Interest" is a term of art that translates to "blogs and editorial and training development efforts" in DMI parlance.
Before you join DMI, you have access to some of our editorial and training content, but to help guide DMI in the R&D directions that would most fit your needs and interests, you need to join the communities that interest you and contribute your viewpoints, experience and ideas.
Consider yourself invited.
DMI's communities are organized like a hierarchical tree. After several decades in the business of operating information technology on behalf of organizations, we have reached the conclusion that success -- regardless of the metrics you employ -- always comes down to how well you manage data.
Data is the irreplaceable asset of an organization. It is the reason for deploying information technology in the first place and managing its storage and accessibility are key determinants of the agility and the success of the business.
Given the primacy of data, it stands to reason that Data Management is the trunk of the tree in the DMI model. Data Management is shorthand, referring to the creation of policies for managing data throughout its useful life, and also for the activities and processes involved in hosting the data on the best storage kit (relative to the value of the data, its update and access frequency, and business rules governing its retention) and for assigning the data the right protection, privacy and preservation services throughout its useful life.
So, data management is the top tier concern of IT, with infrastructure/storage architecture and services such as data protection, data privacy and data preservation comprising branches of the tree. From this perspective, the central or mission of IT should be data stewardship.
Our C-4 Strategy (standing for cost containment, compliance, continuity and carbon footprint reduction) is another way to visualize the DMI tree.
To get to cost-containment in IT, you need to managed data across storage infrastructure wisely and effectively.
Compliance requires that data be managed in accordance with deliberate rules and standards, many of which are required by regulatory or legal mandates.
Continuity is a primary concern because data itself is irreplaceable. The only way to safeguard data is to make a copy and store it far enough away from the original data so that it is not consumed by the same calamity.
And carbon footprint reduction refers to the need to reduce the consumption of fossil fuel generated electricity when operating IT infrastructure. This is undertaken as a cost-savings measure, but it should also be undertaken to reduce the rate of carbon contamination of the Earth's atmosphere, stalling catastrophic climate change.
The DMI tree represents a philosophical view of what IT should be. It may be noteworthy to some that we haven't mentioned virtualization technologies or clouds. We see such technologies as subordinate to the primary goal of data management. In fact, we worry that some of the hyperbole around hypervisor computing and clouds has reduced the skills and knowledge of IT workers.
Server virtualization has a case to make from the standpoint of equipment consolidation and reduced energy expense. However, some hypervisor vendors have sought to use virtualization to deconstruct storage in a manner that simply establishes their hypervisor kit as the controlling entity within IT. Silos of technology and data are the result, which can only lead to increased costs and reduced efficiency in data management.
As for clouds, DMI views the cloud as an airy marketing term with little empirical value. At essence, cloud is another form of outsourcing -- not very different from service bureau computing in the 1980s, or ASPs/SSPs in the late 1990s. Such outsourcing memes draw power from economic hard times and tend to fizzle out as economic conditions improve. However, the current cloud meme has shown remarkable resiliency, possibly because it is so intertwined with the server virtualization and software-defined movements that portend to deliver IT services with greater efficiency and lower cost.
DMI is not hostile to either virtualization or cloud, but we are cautious not to embrace "new" models until they have delivered on their value case. As of this writing, significant work remains to be done to enable clouds and their virtualized/software-defined building blocks to deliver their promised benefits to organizations. Until that happens, virtual machines are just another kind of workload and the same rules apply to managing the data from these workloads as applied to applications loaded to traditional operating systems of the past several decades.
Of course, data management and storage infrastructure decision-making go hand in hand. They always have. Since the early days, we implemented a "self destructive file system" in computing because storage space was an expensive commodity in extremely short supply (so, it made no sense to store versions of files). As a practical matter, storage capabilities and costs continue to set some boundaries on what we can accomplish with data management. What is key, however, is to deploy storage infrastructure that can be managed efficiently. Storage management and administration are up to 6X more expensive than the annualized cost of storage kit acquisition and deployment, and that's a lot.
Storage capacity is a big issue, and one that is getting bigger as the volumes of data created by users and organizations annually tops 60 zettabytes in 2020. Currently, the combined output of flash and disk storage manufacturers amounts to less than 1.25 ZBs of total capacity. That is significantly less than what is needed to store 60 ZBs of new data.
To avoid what we call a "Zettabyte Apocalypse," DMI contends that an "all of the above" storage strategy will be needed. That includes tape technology -- and a lot of it.
Tape has considerable runway ahead, with demonstrations by IBM, Fujifilm and Sony suggesting that massive capacity improvements will be forthcoming for the next decade -- possibly providing the capacity required to forestall the Zettabyte Apocalypse. For awhile, at least.
DMI has been created to support you, the data manager. As someone who produces, stores, modifies, secures, backs up, archives and deletes data assets, you need all of the advise on technologies and best practices that you can get -- especially now. DMI is driven by a passionate desire to advance objectively true information about storage and data management in the face of considerable disinformation and misinformation we too often encounter in the field or the trade press.
Not long ago, we delivered a live seminar/workshop on data protection in Europe. Following the talk, a trainee approached us and stated that he needed to learn a lot more about a technology we kept talking about, something we called "tape." The fellow had never heard of such a technology. It had not been mentioned once in the formal training (and only training) he had received to support his role as a "virtualization administrator." The hypervisor vendor who had trained him had no knowledge of nor use for tape technology in their kit, so the topic was never broached. The fellow said he wanted to learn more about it, since it seemed to potentially solve many problems in his environment, and he wanted us to suggest some additional resources.
He also asked why you needed to build a room full of books to operate a tape system. When we asked for clarification, he noted that we had been talking about a "tape library." We started work on a new course on storage foundations the very next week.
So, welcome to DMI's communities of interest. We hope you will find a topic or area of research that appeals to you and that you will contribute your knowledge and insights so we can all improve our collective professionalism.