Thought Leadership
In times of climate crisis, companies want and need to operate more sustainably in order to reduce their carbon footprint. To achieve this, most organizations are actively working on sustainability strategies or already have them in place.
As part of this mission, organizations inevitably come across one of the biggest energy consumers in many companies: data storage.
Sustainable data storage: easier said than done.
Many companies are looking for ways to store their data more sustainably. However, this is easier said than done, as data storage is extremely complex. The main reason for this complexity is the almost unimaginable amount of newly generated data. The largest proportion of this, around 90 percent, is unstructured data. In addition, companies have stored almost every bit of data they have ever generated and have continuously adapted their storage infrastructure to do so. As a result, the storage infrastructure in most companies is a veritable patchwork of different storage generations, hardware providers and storage technologies, which has become increasingly networked and complex.
The management of unstructured data has many variables.
Many of these heterogeneous storage infrastructures have grown over decades. In addition to the often numerous physical storage units in the data center, there are often several data storage units for numerous use cases in different clouds. Such storage environments are widespread and the operating organizations are constantly faced with the task of finding the ideal mix for their storage requirements. Until now, sustainability has often played a subordinate role. In most organizations, it was much more important to make the right data available to users in the right place at the right time. And this still applies. However, sustainability is now another important variable that needs to be included in the equation of this “sweet spot”.
Data availability is becoming increasingly important.
Much of this data is now business-critical in a variety of use cases, for example when training AI models. This means that the data cannot simply be moved to an inexpensive archive as in the past, but must remain usable. As a result, a new paradigm has emerged in which data must be continuously exchanged between systems in order to maintain business operations. For example, the training data of an AI model cannot simply be moved deep into a long-term archive as it used to be, from where it takes a long time to move the data back to an active primary storage. Developers require the data to be available at all times in order to continue training the model, improve it and add new functions. These new challenges require a much more dynamic management of unstructured data than was previously the case.
Overview of the CO₂ balance of unstructured data.
Infrastructure and the management of unstructured data are prime examples of the influence that sustainability can have on the technology used. If you compare how much energy different storage systems consume, you can see that there is a wide range in terms of energy consumption: Flash storage consumes less than hard drives and cloud storage is usually much more efficient than on-premises due to scaling and the location of cloud data centers. Understanding how much energy each storage location and type consumes helps companies to determine the appropriate storage for each type of data and thus save energy and CO₂. To achieve this, companies should be able to monitor their key indicators and take appropriate action. Active management of unstructured data enables companies to CO₂-balance unstructured data storage in the cloud and in the data center.
Measures for more sustainable data storage.
With a clear overview of the company’s carbon footprint, IT managers can, for example, identify which unstructured data should be moved to the cloud and what measures need to be taken to optimize the data center. These measures include deleting redundant, outdated and trivial data (ROT) or moving data to the cloud to less environmentally harmful storage. Or the consolidation, reuse or decommissioning of hardware. Obsolete and inefficient storage can be identified and decommissioned by migrating data to unused storage space. The purchase of duplicate backup hardware in each branch office can also be avoided and a backup performed in the main data center instead.
Conclusion: Platforms for the management of unstructured data improve sustainability.
Climate change is no longer just a distant threat for business leaders. For good reason, sustainability is now at the top of the agenda of almost all large companies. The first and most important step in storing data more sustainably is the active and planned management of unstructured data. Without proper management of this data, organizations will face a number of critical issues, including their ability to comply with sustainability policies. To do this, companies need two things above all: firstly, accurate insight into the carbon footprint of their stored data and tools to manage the data so that it is stored in the optimal location. Modern, vendor-neutral multi-vendor and multi-cloud platforms for the management of unstructured data are suitable for this.
