Cross-Platform Storage Automation Enables Greater Value from Data
Introduction to the problem
There is a long-held dictum that complexity drives cost in IT. That is true of storage infrastructure in particular, especially as all industries are struggling with how to reconcile the cost of storing, protecting, and managing increasing volumes of unstructured data.
At issue is the fact that just as there are many types of data — each with different value to a company at a given point in time — there are also many storage choices available to those companies, each of which is optimized to only a portion of the data requirements at that moment. Data that is seldom used, for example, doesn’t need to be stored in the same way that active data are. Some subsets of data may need very high-performance all-flash arrays for a period of time, whereas others may work well with object storage, tape, or cloud. Multiple copies may need to be managed for DR, collaboration or other uses.
The problem is that there is not a “one-size-fits-all” storage solution that addresses all of these data requirements, particularly without interrupting users, or adding management overhead to IT. Even with years of advancements in storage technologies, the management of storage infrastructures across multiple vendors platforms has only gotten more complex especially as IT organizations have to do more with static or shrinking operational budgets.
Storage vendors try to address this by offering end-to-end storage-centric “ecosystems”. This creates a vendor “silo” that locks users into a particular file system or the vendor’s solution stack. This type of solution might work fine for users that can commit to a single, one-vendor solution for all their storage needs. But such vendor-locked solutions typically come with a premium price tag, and significantly limit customers’ freedom to take advantage of other storage technologies that may emerge. But most users have several different file systems, and multiple vendor solutions which don’t work together. Too often IT organizations are left with no choice but integrate their own mix of storage solutions, which adds risk, complexity, and significantly adds to operational expense.
What needs to be done is simple: IT planners need to find ways to drive complexity out of storage and to automate storage operations. That is, to simplify and automate data management across heterogeneous storage types, and of course improve the ROI of the entire storage infrastructure.
Operational costs typically drive over 80% of annualized storage TCO.
Cost of ownership and ROI
These days, IT departments are finding their data storage strategies subjected to the same rigorous accounting scrutiny as are the non-IT initiatives of other business units. Concerns about total cost of ownership (TCO) and return on investment (ROI) are pressuring IT planners to contextualize their funding requests with utilization and other empirical metrics.
From the perspective of TCO, there is good reason to put storage in the crosshairs. Depending on the analyst you want to reference, storage accounts for between 33 and 70 cents of every dollar spent on IT hardware annually. Moreover, analysts estimate that the demand for storage capacity of unstructured data is growing at a rate of more than 200% per year, than the 40% annual rate that was estimated in 2010. This increase is attributed to a variety of trends, but to storage administrators, this translates into needing to spread existing resources across an ever-growing data management problem.
Not surprisingly, the cost trajectory of storage acquisitions has attracted significant attention. Tactical stop-gap technologies like compression and de-duplication have been enlisted to squeeze the maximum amount of data into the minimum amount of storage space. However, these techniques do not address the real source of storage expense, which is only partially explained by equipment and software costs.
In fact, the cost to acquire and deploy storage is only a fraction of the cost to own and operate complex storage infrastructure. By some estimates, annualized storage TCO has a 6:1 cost ratio – with the cost to acquire, license and deploy representing only 1/6th of the total cost to own and operate storage infrastructure (calculated on an annual basis). The other 5/6ths of the cost to own the infrastructure consists of management and administration (labor costs), backup and archive, facility costs including floor space, power and cooling, and other day-to-day operational expenses.
The increasing number of contemporary storage choices, whether monolithic or software-defined, actually makes administration and management more challenging, not less. Greater complexity in storage arrays requires administrators to possess greater skills and experience, which costs more money. Integrating disparate storage solutions into unified topologies and moving data between them requires manual intervention, scripting, and other workarounds. And every maintenance upgrade or technology refresh, even minor ones, carries with it a risk of downtime, disruption to users and more hands-on effort by IT staff.
The reality is that real storage TCO is usually determined more by operational expenses (OPEX) — administration and management, facility costs, data management costs — than by capital expenses (CAPEX) – acquisition, licensing and deployment costs. This fact is inescapable and requires attention to do whatever we can to drive down OPEX in storage.
At issue is the complexity of the solution. Complex storage kit requires more attention by storage admins or greater dependency on the vendor. Complex value-added functionality, whether deployed on an array controller or behind a particular hypervisor, can actually obfuscate the global management of infrastructure and data, driving more manual processes. Complex, manual data management processes generally require more operators, driving OPEX higher. And, complex infrastructure, administered by more operators, is generally more prone to expensive downtime, which requires more complexity in the form of data protection and preservation.
Object-based storage and cloud solutions are often mentioned at this point as a savior of storage TCO, but for companies with large volumes of data under management, cloud storage alone is not the answer. Larger enterprises have been careful when embracing clouds as a replacement for their on-premises data centers. Many have opted to leverage clouds for specific applications or as extensions to their infrastructure. Extending an already complex infrastructure to include clouds often increases, rather than decreases, complexity and TCO.
The reality is that most IT organizations simply have no way to measure how efficiently their most expensive storage resources are being utilized across the enterprise.
Bottom line: TCO challenges remain front and center even in organizations that have embraced cloud and software-defined storage architecture. Lousy TCO, in turn, has a tendency to sour the moods of budget hawks who want to see measurable returns on storage technology investments.
Achieving a measurable improvement in ROI
Storage infrastructure and data management are cost centers to every organization, regardless of the data type or use case. Even — or perhaps especially — when data is the company’s primary asset, the ROI for how those assets are stored and protected directly impacts the bottom line.
This goes beyond the way ROI is often calculated on storage vendor web sites, which may only measure how an individual storage element compares with an alternative solution. What is typically omitted from such ROI calculations is the impact complexity for IT staff to manage another storage type, or disruption to users when their data moves from one platform to another. Looking at such measurements in isolation of the whole picture is a storage-centric attempt to answer data-centric problems.
The oft-quoted business maxim that “you can’t improve what you can’t measure” applies directly to this situation. The reality is that most IT organizations simply have no way to measure how efficiently their most expensive storage resources are being utilized across the enterprise. It is easy enough to know how fast the data is growing in each storage tier, and when more storage will need to be added. But with most storage tools today, it is very difficult if not impossible to provide detailed real time metrics of storage utilization across all storage types, active vs passive data under management, project or departmental storage cost allocations, and other information necessary to optimize data placement across storage resources, and thus to maximize storage utilization.
Truth be told, the best place to start simplifying and automating storage is at the process layer, not the device or infrastructure layer.
Automating cross-platform management
StrongLink is a software solution to provide IT organizations data management automation and the data insight across all storage types, from any vendor. StrongLink addresses the problem of bridging multiple storage silos by leveraging metadata derived intelligence combined with storage resource management tools to provide end-users complete control over their data and their storage.
StrongLink puts customers in control, providing independence from vendor silos, or problems such as globally managing multiple storage use cases.
Bridging otherwise incompatible file systems and storage types, StrongLink maximizes storage utilization, automates data migration, tiering and protection. StrongLink empowers customers with visibility into their data, and real-time reporting and data insights as well as the telemetry to improve workflow.
StrongLink puts customers in control, providing independence from vendor limitations, proprietary applications, or the problems related to managing multiple storage protocols and use cases. Data is presented in a global namespace and file system view, users only see their data based on permissions mapped from their AD credentials. This means users can simply focus on using their data, and stop wasting cycles trying to find it or retrieve it.
StrongLink removes complexity even in heterogeneous storage infrastructures. Driven by metadata and easily defined policies, data life cycle management is executed dynamically to support any workflow while ensuring data is always protected with data integrity verification.
StrongLink maximizes storage utilization efficiency by enabling the proper placement of data on the proper elements of storage infrastructure based on policy. Placement of the data, both initially at the time of data creation, and over the useful life of the data, is handled by cognitive processes that evaluate the state of the data, the status and availability of infrastructure, and the status and availability of storage services all within the context of a predefined policy for the data. It dramatically reduces the number of operators required to manage the infrastructure while ensuring that you get the most from the capacity that you have. It allows IT administrators to automate data polices across storage tiers and locations. This includes the ability to introduce new storage tiers or perform seamless migrations and tech refresh without interrupting user access.
StrongLink makes storage scaling significantly simpler. Whether you wish to expand the capacity of an array or add a different type of storage to your infrastructure, StrongLink is able to seamlessly incorporate the new capacity into its managed pools, adjusting the distribution of data dynamically and in accordance with data management policies. That saves CAPEX by providing planners with the flexibility to purchase the storage they need without concern for brand name, and it saves OPEX by automating tasks that were once performed manually, such as re-distributing data across new capacity.
StrongLink provides data protection, preservation, and privacy services via its own stack, but it can also work with third party backup, archive, or security software that you may currently be using. Ultimately, StrongLink associates data with the “best fit” protection, preservation and privacy services available in your infrastructure automatically, according to policies and priorities you decide. And, StrongLink’s cognitive capabilities ensure that no single source of services becomes overloaded by too much demand.
Getting to a meaningful ROI improvement in storage and data management requires relentless simplification and automation. But it must be business savvy. What’s needed is a cognitive management technology that will extend to all storage – on premises and in the cloud – facilitating ease of administration, optimal capacity allocation and utilization efficiency, and rapid, agile provisioning with metrics in response to business needs.