Data Analytics in the Cloud: It's Not Just Lift and Shift

Today, adopting the cloud is a foregone conclusion for the enterprise. As the scope and diversity of data grows at an increasingly faster pace, it’s no wonder that nearly 60 % of North American enterprises now rely on public cloud platforms, five times the percentage that did just five years ago[1]. Gartner predicts that by 2022, 75% of all databases will be deployed or migrated to a cloud platform, with only 5% ever considered for repatriation to on-premises[2]. It’s clear that the cloud’s flexibility will rapidly become an essential success factor for businesses across every industry.

But what’s become lost in many industry conversations about the cloud is the fact that the most successful companies don’t just migrate their existing systems, or “lift and shift.” They design an entirely new architecture built for the cloud, one that allows them to completely rethink how they run their businesses. This isn’t about transferring their current operations onto a new form factor; this is about creating a new, powerful platform that allows them to keep up with the speed and diversity of data growth, the pace of innovation and the unique needs of the enterprise.

This is about a whole new way of thinking, where enterprises design and innovate with the cloud in mind. This new mindset is about being “cloud forward,” whether or not you were “cloud first” or “cloud native” from the start. Any enterprise can adopt this mentality, no matter where they are in the cloud journey.

The challenges of designing a cloud architecture

Too often, companies attempting to retool themselves into cloud-forward businesses make some common errors that hold them back in their transformation:

Lack of provider and deployment choices.

Some companies choose a single cloud provider, or they adopt a solution that doesn’t support multiple deployment options. This causes them to be locked into certain providers’ terms and offerings, making it difficult to rebalance their services in response to future business needs.

Compute and storage that only scales together.

The enterprise may also encounter challenges if they adopt cloud solutions that scale storage and compute solutions simultaneously. Companies do this in order to manage massive amounts of data, thinking that they’ll maximize performance while saving costs. But it’s important to be able to scale these capabilities independently. If data is streaming in at a rapid rate, having to scale compute nodes at the same level as storage could drive up costs when the value of the data is unknown. Or, demand might drive the need to scale up compute based on what the business user needs even as the amount of data being stored doesn’t change.

Expensive and inflexible data storage.

In recent years, companies have chosen to store data on Hadoop in the hopes that this would be a cheaper and more flexible option. However, many enterprises are finding that Hadoop offers the complete opposite. It limits them to process data on-premises. And while Hadoop makes it easy to transfer data into an enterprise’s system, it’s difficult to get that data out when it’s needed.

Inflexible pricing.

Some enterprises adopt cloud solutions that charge them to “rent” tiered levels of processing space and services. Others sign contracts that locks them into a year or more of terms and payments. When enterprises can’t simply pay for the compute and storage they use, they can’t fully benefit from the flexibility that the cloud promises, especially when the needs of their businesses are likely to fluctuate month to month. Some companies also attempt to design their cloud architecture by connecting various point solutions from multiple vendors. While these may be powerful and innovative tools, the enterprise doesn’t always integrate them in harmonious ways that maximize their potential and usability.

Lack of core production.

Finally, some companies may quickly design a cloud solution that they only use for testing and developing new products. But without a core analytical production capability, these companies lack the foundation they need to fully innovate.
 
In my next post, I’ll discuss how Teradata Vantage addresses these challenges to empower enterprises to take full advantage of the best data analytics with all the benefits of the cloud.
 

---