After providing a viable alternative for enterprise IT systems for nearly two decades, are NoSQL databases finally making significant inroads against the conventional proprietary world of SQL parallel databases?
If you look at the marketshare numbers, this is true. New-gen providers such as MongoDB, MariaDB, MapR Technologies, DynamoDB and MarkLogic are leaders in the big data NoSQL space, according to Forrester Research, and they are closing in on the market dominance of the Oracles, SAPs, Teradatas and IBMs of the world.
This was made quite evident recently when Amazon Web Services, the most powerful cloud services provider on the planet, revealed that it is in the process of weaning itself off Oracle and moving to alternative databases—such as its own DynamoDB and others--to carry its humongous daily workload. AWS has said it wants to make the move permanent by 2020.
AWS in Process of Making a Big Move
There’s no question that Oracle has been a key reason why AWS has ascended to global IT heights in the first place. However, as AWS has scaled out, it now perceives a need for new-gen data storage inside DBs that are easier to manage, not as expensive to maintain, and more flexible in integrating and moving workloads.
NoSQL technology offers enterprises flexibility because NoSQL data stores can support structured, unstructured and semi-structured data for different types of business applications. Older SQL databases have issues with scripting languages, such as JSON, for example, and are more limited in scope than the newer ones.
Forrester has cited MongoDB as the most popular NoSQL database for the last couple of years. The open-source database is "popular among developers because it is easy to use, scales to meet the most demanding applications and offers the most comprehensive ecosystem of tools and partners," the researcher said.
MongoDB boasts more than 2,000 paying customers using its open-source database, including companies such as AXA, Barclays, Comcast, DHL, Expedia, Sprinklr, UPS i-parcel, The Washington Post and x.ai.
NoSQL Vendors Making Major Inroads
But others, like San Carlos, Calif.-based MarkLogic, are also growing (more than 4,000 customers) and making numerous converts. MarkLogic was the NoSQL database that helped solve the online registration snafu of Healthcare.gov, the Affordable Care Act website, in 2014. Since then, it has become a word-of-mouth convincer of government and military agencies to move from old-school to new-gen databases.
It keeps evolving, too; the company recently launched the MarkLogic Query Service, a new way to give customers automatic elasticity in the cloud for their enterprise workloads.
The new software-as-a-service application automatically adds and removes query processing capacity to a database cluster as workload demand rises and falls, resulting in agility and responsiveness for end users. This has the potential save a lot of time and effort for IT managers as they schedule various workloads on premises or in single or multiple clouds.
It’s because of this versatility of working with any kind of workloads, on-premises or in any type of cloud, that makes databases like these attractive to IT managers. Data comes into a system in all kinds of forms, and all these forms need to be accommodated, whether they are structured, unstructured, block, file or anything else.
Why There Is Major Disruption
This is all causing a major disruption in the way companies are buying databases here in the second half of 2018.
“The disruption we’re seeing is showing up in the verticals that are embracing it, in the geographicals—in terms of companies pushing it outside of the U.S.—and it’s expanding relative to the depth of what our relationships have become within the accounts we’re in,” Gary Bloom, CEO of MarkLogic and a former Oracle executive, told eWEEK.
“Basically, anytime anything changes in a conventional relational database, the ETIL breaks [extract, transform, load--three database functions combined into one tool to pull data out of one database and place it in a data warehouse]. This happens all the time in the real world. Standard relational databases are anything but versatile.”
This kind of disruption in a workload is both expensive and time-consuming. Lighter, more agile NoSQL systems are filling the bill for a boatload of use cases, including anything transactional, operational or analytical, Bloom said. Verticals making the switch include fintech (which tends to move data workloads frequently), government, science, health care, retail and media.
This is a trend that has legs, according to a majority of IT analysts. We’re keeping a close watch on this here at eWEEK, because it’s fundamental to how IT performs.
Here are the market leaders overall in enterprise database marketshare in 2018 (in rough sales-leader order):
Oracle 12C, Microsoft SQL Server, SAP Sybase ASE, PostGresSQL, MongoDB, MariaDB Enterprise, MySQL, Teradata, Informix (IBM), Ingres, Amazon Simple DB, MarkLogic, Couchbase.