This post first was written for the Forbes Technology Council by Krishna Subramanian here.
The cloud is increasingly seen as a key IT strategy for businesses, as over 90% of enterprises are shifting to a multi-cloud strategy, according to a study by The 451 Group. Not only are businesses adopting the cloud more, they are also increasing their use of cloud resources by shifting from a “cloud transformation” strategy to a “cloud-first” strategy.
As the President & COO of a data management company that focuses on cloud storage, I see this shift creating a gold rush to support core enterprise workloads in the cloud.
Cloud Transformation vs. Cloud-First
What is the difference between cloud transformation and cloud-first? For the past few years, enterprises have been on a cloud transformation journey — a measured approach to grow cloud adoption for the appropriate use cases. Often, this meant that cloud-native applications — new applications written to take advantage of the cloud — were the ones that moved first. Two good examples of cloud-native applications that have seen tremendous growth are Software-as-a-Service (SaaS) and Developers and Operations (DevOps) teams using the cloud for testing and development. In fact, Forrester Research proclaimed 2017 as “the year of enterprise DevOps” as the DevOps market hit escape velocity.
Businesses are now more familiar with cloud benefits from these early forays and are moving from a cautious cloud transformation approach to a more bullish cloud-first approach. The pandemic has also accelerated this shift, as data centers have been harder to access and the workforce has gone remote.
A cloud-first strategy is fundamentally different from a gradual cloud transformation approach because it means enterprises will look to add any new infrastructure in the cloud first — even for existing enterprise workloads — before they add any datacenter capacity. This means traditional enterprise workloads that have been running data centers will now have to run in the cloud.
This is a pretty significant shift.
The Rise Of Cloud Network Attached Storage
The data formats used by enterprise applications are different from what the cloud traditionally supported, and so enterprise applications had to be written as “cloud-native” to run properly in the cloud. This is a big reason why traditional enterprise applications have been slow to move to the cloud until now. A recent survey found that 77% of enterprises are looking for more data integration and workload mobility across clouds.
Most enterprise applications deal with unstructured data. Today, according to CIO, over 90% of the world’s data is unstructured — genomics, self-driving cars, audio, video, seismic data and documents are some examples of unstructured data. Most unstructured data used by traditional enterprise applications is stored as files in storage known as Network Attached Storage (NAS). NAS environments are designed to deliver high performance, but they are also expensive. The cloud supported a cheaper way of storing unstructured data as objects, not files. This is usually significantly less than the cost of file storage, but traditional file-based enterprise applications cannot run on it.
But that is now changing. All the major cloud providers and enterprise file storage vendors are currently rushing to deliver cloud NAS or file storage in the cloud. Similar to NAS in a data center, this cloud NAS is designed to support the performance needs of enterprise workloads, but it is also significantly more expensive. As an example, Amazon Web Services, the largest public cloud provider, has EFS file storage ten times more expensive than their S3 Object storage. With such a huge difference in costs, the new file storage in the cloud should only be used when needed.
Better Cloud Management Emerges
Managing cloud costs is already a challenge for enterprises. In fact, Gartner Research estimates that in 2020, 80% of enterprises will overshoot their cloud costs. A new category of cloud data management software is helping enterprises address this issue by using analytics to understand data usage patterns and automatically move data in the cloud across file and object storage to optimize both costs and performance.
Data management can also help businesses unlock the value of their data by creating virtual data lakes with just the right data needed for new applications like artificial intelligence and big data. Good data management is about having the right data in the right place at the right time, especially because the majority of data becomes inactive or cold within a year of creation yet consumes expensive storage and backup resources.
Data management ensures that actively used hot data gets the best performance, albeit at a higher cost, and that inactive cold data is still accessible, but costs much less to store. This involves using data analytics to understand how data is being used by the business and then moving data based on analytics. This is done through techniques such as data migration, data archiving and tiering to dynamically move less-used data to cheaper storage classes, and global search and tagging to still find data easily no matter where it lives.
When looking for a data management solution for your company, there are three key things to consider.
- First, how do you expect your data strategy to evolve over the next five to ten years? If you want the flexibility to change your cloud or storage providers, then consider a data management solution that works across a variety of storage and cloud platforms, but if you don’t need this option, you can consider a vendor-proprietary solution.
- Second, think about your budgets and the skillsets of the people who will be managing data — many customers are choosing data-management-as-a-service solutions because they are easy to deploy and use.
- Third, consider your data control posture: Data is the new oil and it’s one of your strongest corporate assets. With that in mind, how important do you think it is to keep control of your data? If it’s truly important, you want to understand if the data management vendor creates any lock-in.
With public cloud spending soaring to $331B by 2022 according to Gartner, the gold rush is on for enterprise workloads in the cloud.