Eliminating Database Sprawl, Part 1: How to Escape a Slow-moving Car Crash


Rick Negrin

VP, Product Management

Eliminating Database Sprawl, Part 1: How to Escape a Slow-moving Car Crash

How do companies veer dangerously close to the perils of database sprawl? Even more importantly, how do they correct their course? Find out for yourself as Rick Negrin, Field CTO at SingleStore recaps the webinar presentation on  this topic by Forrester vice president and principal analyst Noel Yuhanna. This is Part 1 of a two-part series on eliminating database sprawl.

Companies often aren’t aware they’re on a collision course with database sprawl, a lumbering jalopy of high-maintenance databases that can T-bone any organization at a busy intersection—for example, during demand  forecasting, when teams attempt to  integrate and rationalize data pulled from multiple systems of record, transactional and analytic databases, relational and no-SQL databases, graph analytics databases  and ad hoc inputs in myriad data formats.

How do companies get caught in database sprawl and, even more importantly, how do they get out, to lower costs and accelerate use cases? This two-part blog recaps SingleStore’s recent webinar, “Eliminating Database Sprawl: Three Customer Case Studies“ featuring Noel Yuhanna, vice president and principal analyst at Forrester, and Rick Negrin, Field CTO at SingleStore.

tightly-coupled-databases-fuel-sprawlTightly coupled databases fuel sprawl

“As an analyst, I get to speak with about three to four customers every day,” Noel said, introducing the topic.  “These are Fortune 1000 companies and mid-sized companies with data platforms and databases built on premises and in the cloud. Now that we're actually in the cloud and multi-cloud, they always run into issues with different databases, such as how do you manage, and better manage them? This becomes a nightmare for a lot of organizations, especially as the data explosion continues and you're building new applications.”

Noel then dove into the backstory of how database sprawl typically occurs, through development of a wide range of siloed applications—web, customer 360 CRM systems, e-commerce, IoT and many more—each with a tightly coupled database and its own compute, storage and security resources. Separate as they are,  data is required to move easily between the siloes.

“Over the last many decades, we’ve built these tightly coupled applications with the application and the database together,” Noel said. “The bigger downside is data movement, when you try to synchronize data between this source and that source, in the cloud and on premises.”

He illustrated the sheer size of database sprawl: “I was speaking to a financial services company and they have more than 50,000 databases. They are now starting to move to the cloud, but most of the databases  are on premises because they contain data that is sensitive or classified. In these environments, whenever you build more applications, the situation becomes that much more complex.”

data-management-challengesData management challenges

Database sprawl creates multiple problems that snowball over time and can strike business to a halt--just like a car crash. Noel spent several minutes reviewing a host of data management-related problems associated with sprawl:

  • Slows down new development
  • Creates security/compliance issues
  • Duplicate data leads to inconsistency
  • Create data integration issues
  • Under- or over-provisioned server waste resources
  • Increases administration complexity
  • Impacts service level agreements (SLAs)
  • Increases cost, impacting budgets

SLA-killing latency bore additional discussion. “We believe more than 30%--that’s three-zero—of data in an organization is duplicated. As data gets copied it creates a bigger challenge, right?” Noel asked rhetorically. “How do you actually ensure consistency and trustworthiness of data? It’s extremely challenging when the applications demand consistent, trusted data and increased data movement concurrent with administrative tasks such as backup and recovery, disaster recovery, database tuning and batch management upgrades.

“This is an important element; simplified database administration is what customers are looking for, using automation and intelligence,” Noel concluded. “Reducing the costs of databases is a big component of what customers are demanding, based on the inquiry calls we get from financial services, retail, manufacturing and many other industries.”

achieving-faster-time-to-valueAchieving faster time to value

He described the essence of the solution that customers want to put  an end  to database sprawl: “Customers are demanding, ‘Hey, we really want to have a faster time to value. We have to deal with multiple data types and data models, and latency has to be minimized. And the cost should be controlled in ways besides by simplifying administration.’”

Noel outlined the capabilities of modern database platforms that can help reduce sprawl. One such platform is what he calls a ‘translytic database’: “This is basically combining transactions and analytics and operational workloads together in a single unified platform,” he said, explaining, “Traditionally we have had transactional systems, separate operational systems and separate analytical systems. There’s never been a combined platform. That slows you down, because by the time data goes through the data loops, and all of the data processing is done, and ETL is done, the dashboard is stale. It's 24 hours late, actually, right? The correct data is not showing up.

“That's what a translytical database solves—it combines these different workloads together in a single platform, to drive and deliver analytics and operational insights at the speed of transactions.”

According to Noel, in addition to translytic capabilities,  platforms to consider include:

  • Multi-model platform that combines multiple data types and models together in a single platform. This encompasses relational, document, graph and others.
  • Hybrid cloud databases that deliver a common platform across on-premises and public cloud, for applications that demand consistent data.
  • Edge databases that deliver data at the network edge, to support modern business applications requiring faster access to data and granular security.

Noel wrapped up his portion of the webinar with specific recommendations on how to do away with database sprawl: “The recommendation we have is to look at a modern database platform to support new and emerging business requirements, a platform that goes beyond the traditional database functions.” His selection criteria include:

  • Look at translytical, multi-model, edge database and hybrid-cloud databases that can help deliver consistent, trusted and real-time data for apps and insights.
  • Consider consolidating database silos to improve administration efficiency, lower cost and future-proof your database
  • Keep security in mind when consolidating databases
  • Focus on eliminating high license cost databases by replacing them  with a modern database platform.

With any luck, Noel’s insights can  help your company to dodge the inevitable crash of database sprawl. Come back next week for part two of this blog series, which dives into the webinar’s customer use cases, “Eliminating Database Sprawl, Part 2: How Three Companies Beat the Odds.” In the meantime, check out the database sprawl webinar, keep up with our latest news on Twitter @SingleStoreDB and follow our new Twitter channel for developers @SingleStoreDevs.