By Martin Gibbons, Channel Director for EMEA at Komprise
Working effectively with information systems in modern cloud-native enterprises across a massively diverse data topography requires a comprehensive approach to information management across the full lifecycle (and retirement-cycle) of data itself can we now isolate some core home truths to organise our way out of information chaos and pandemonium?
Modern enterprise IT systems are rarely design-perfect structures fit for the beauty pageant or best-in-show rosette in the annual information architecture awards, were such a ceremony or trophy to exist.
Most enterprises operate atop a tangled spaghetti structure of applications and technologies, all sporting different data streams at varying degrees of structured, semi-structured or wholly unstructured preparedness. Data is everywhere, and too often hidden or underutilised and certainly under-managed as silos have grown with the cloud. So, given the virtual ‘information pandemonium’ that is playing out, how
can we cut through the white noise?
Data outlives storage, storage dies first
Let’s start at a somewhat terminal point of reference to clarify the issue. Today we know that unstructured data growth and storage proliferation are urgent problems that IT organisations can no longer ignore.
When a business can grasp the steps, stages, levels, and tiers that data resides at in this way, it can maximise data management savings with file-level tiering. Less actively used data can be tiered to less expensive storage and all data has a richer degree of meta-level intelligence denoting which users have accessed it when and where and how often.
By understanding enterprise data across the entire environment, IT decision-makers can maximise savings with a more intelligent approach to the storage and backup cycles. The storage savings can be significant, but even this advantage is generally dwarfed by the savings achieved through eliminating the existence of cold files from recurring backups.
But data outlives storage, in that the life, value and usefulness of data will almost always extend beyond the life and value of any storage technology, cloud-based or otherwise. IT directors need the flexibility to move their data to new devices and clouds as is desired and required to meet business requirements and achieve cost, performance and data protection advantages. They want to do this without experiencing the penalties (in the form of license fees, integration hassles or cloud egress when data leaves a private network to reside on some other service etc.) associated with tiering, archiving or migrating data elsewhere.
The point is to avoid the lock-in that comes from buying and renewing expensive legacy storage platforms, which, in practice, are typically very difficult to fully migrate away from onto newer technologies and cloud services.
Breaking the silo mindset
This uncomfortable truth should be a wake-up call to any organisation that still insists upon managing data through any form or shape of storage silo. IT departments should instead embrace and implement data-centric management that exists as a separate layer working across all storage and clouds to analyse and move data without creating lock-in. Rather than trying to get rid of silos—a hopeless effort in multi-cloud and hybrid IT environments—IT leaders should determine how to work across them. Storage vendors have data management features, yet those tools only work with data stored on that platform, perpetuating the silo conundrum.
Independent data management provides the visibility required to gain a holistic understanding of unstructured data assets: where data lives, its costs to store and backup, and its key attributes such as owner, file type, age and access patterns. This knowledge gives IT the means to make storage decisions that maximise the value of data, optimise costs and increase agility.
Extending the life and ROI of data
IT organizations should look beyond cost savings with data management. It’s time to channel data repositories and services into new process and use cases with cloud-based analytics tools which can generate new value.
Cloud is enhancing the power of AI and machine learning – of which organisations like Snowflake and Databricks have already taken full advantage. New data analytics platforms can help business leaders make use of their unstructured data for competitive gain or to meet other business goals, such as sustainability regulations or customer satisfaction.
While it’s been previously challenging to use unstructured data, there are now ways to do so effectively. And it’s getting even easier with cheaper storage and tools for citizen development that doesn’t require users to have a PhD.
Data management and storage vendors play an important role in helping manage and move data to places where it can be leverage and protected for future AI and machine learning projects. By placing data correctly, organisations will waste less money on exorbitant data centre storage and can move resources to cheaper alternatives. This money saved can then be piled back into R&D tools that make more efficient use of the data.
A data management platform can deliver a granular understanding of the differentiation between real-time data, current transactional and operational data, cold or stale historical data and legacy data that needs to be retired or deleted.
The end result is data management with visibility and precise search across multiple storage systems and clouds. Business users can more cost effectively perform analytics and gain actionable insights on the data types that exist across the entire data estate.
Human evolution & data revolution
Looking ahead, we can say with some certainty that our enterprise data responsibility is only set to widen, broaden and become increasingly complex. We know that more data has been created in the last two years than in the previous whole history of humanity.
Given that truth, imagine what the next half-decade will produce. With quantum computing just round the corner, and Artificial Intelligence (AI) and Machine Learning (ML) continuing to fuel data creation, dynamic data proliferation must be matched with an intelligent, automated approach to manage it.
Pandemonium roughly translates through the ancient Greek as ‘all demons’… so if we now must manage ‘all data’, then we will have to hope for divine intervention or sublime data management… and we all know which is the safer bet.