michelangelus - Fotolia
Open source RDBMS uses spurred by lower costs, cloud options
Open source technologies have become more capable alternatives to mainstream relational databases for cost-conscious users -- and the cloud makes them more accessible.
As the volumes of data generated by organizations get larger and larger, data professionals face a dilemma: Must database bills get bigger in the process? And, increasingly, IT shops with an eye on costs are looking to open source RDBMS platforms as a potential alternative to proprietary relational database technologies.
Growing support for open source relational database management system software among cloud computing vendors provides another impetus for users to explore tapping it to replace the likes of Oracle and SQL Server, according to Rick Sherman, managing partner at consulting firm Athena IT Solutions.
"The open source databases have been around for a while," Sherman said. "Especially on the web side, developers have used MySQL, PostgreSQL and similar things. But now we're seeing more open source usage driven by the fact that cloud vendors such as AWS, Google and Microsoft are providing those systems."
As a result, enterprise users "can avoid Oracle and SQL Server fees" without having to deploy a new systems infrastructure to run an open source database themselves, added Sherman, who also teaches business intelligence and data warehousing classes at Northeastern University's Graduate School of Engineering.
Less resistance to open source software
Guy Harrison, the author of Next Generation Databases: NoSQL, NewSQL and Big Data, agreed that cloud-based versions of open source RDBMS products have reached a tipping point in adoption, driven largely by cloud platform market leader AWS. "Amazon has created a business model in which they essentially provide certified versions of open source software," said Harrison, CTO at Southbank Software, a vendor of database development and administration tools. "This has removed one of the resistance points for open source databases -- namely, 'Who do I blame if things go wrong?'"
Further illustrating the pull of the cloud, Gartner pointed to AWS as the revenue leader for open source database software overall, including relational and NoSQL systems.
AWS has plenty of company in the market, though. Microsoft and Google both offer MySQL and PostgreSQL managed services on their cloud platforms, and Microsoft is working to add support for MariaDB, a compatible fork of MySQL. Other open source RDBMS competitors include Oracle itself, which owns the MySQL technology, thanks to its 2010 acquisition of Sun Microsystems; MariaDB, the driving force behind its namesake software; and PostgreSQL vendor EnterpriseDB Corp.
The open source technologies and services offered by those vendors don't come without cost, Harrison noted. For example, users have to pay for technical support. "While open source databases are cheaper [than conventional products], they aren't free," he said.
However, in a research report published in February and updated in May, Gartner analysts Merv Adrian and Donald Feinberg outlined the potential cost savings of deploying an open source RDBMS. Comparing list prices for a sample system configuration, they said MySQL, MariaDB and EnterpriseDB's EDB Postgres Enterprise database would all cost significantly less than Oracle Database on licensing and support fees over the course of three years (see "Database cost comparison: Oracle vs. open source alternatives").
Open source database migration path
Lower costs and cloud availability were both part of MariaDB's appeal as an Oracle replacement to Financial Network Inc., said William Wood, director of database architecture at FNI, which sells software that manages loan and credit origination processes for lenders, retailers and other corporate customers.
Deploying the open source RDBMS was also something of a homecoming party for Wood. While his previous professional work mostly involved Oracle databases, he cut his teeth on MySQL as an undergraduate. "Back in my university days, I did a lot of work with MySQL and Perl, and I knew the power of open source [software]," he said.
When Wood started at FNI in 2010, the St. Louis company's database architecture was entirely Oracle-based, partly because it was imperative to secure data via encryption. Eventually, open source database vendors added encryption capabilities to their products. Wood said FNI began switching the systems that process credit ratings for separate customers to MariaDB in 2016.
Some systems with particularly complex Oracle logic built into them may never be fully converted to MariaDB, according to Wood. But the open source technology has become FNI's primary database platform. "We're basing our future on MariaDB," Wood acknowledged.
William Wooddirector of database architecture, Financial Network Inc.
The motivation for the move isn't fueled by outright cost savings only; it's also driven by a desire to alleviate the overhead work involved in managing commercial database software licenses. In the Oracle world, using a utility like encryption adds to the cost and requires database administrators to track its use on every processor that runs the database software, Wood said.
"With Oracle, you pay extra for every little thing," he explained. "But if you want to use encryption with MariaDB, you just do it. You don't have to go out and count CPUs. The fact is you can spend a lot of time managing your licenses with Oracle. It's not simple."
The migration process has been relatively easy, he added. "MariaDB is almost a drop-in replacement for Oracle," said Wood, who has written a book titled Migrating to MariaDB, due to be published later this year.
Widespread use of the cloud is still in the future for FNI, although Wood said the company does cloud-based disaster recovery through a combination of backups and live replication to a MariaDB system. That's also meant as a proof-of-concept for a possible full transition to the cloud, he noted.
New features and functionality
Enhancements to open source RDBMS platforms continue apace. For example, last month, EnterpriseDB launched a beta version of a tool designed to streamline migrations from Oracle databases to the EDB Postgres Platform. That followed a 10.0 release in late 2017 with improved Oracle compatibility, partitioning and parallel querying for analytical uses.
A version 3.0 update to the MariaDB TX software in May likewise added Oracle compatibility features, plus support for temporal data processing. And MySQL 8.0, which became generally available in April, included performance enhancements for read/write and I/O-bound workloads, default authentication settings and support for new JSON functions.
Another sign of the growing force behind open source RDBMS technologies is their standing in the database popularity rankings calculated monthly by the DB-Engines website. MySQL and PostgreSQL were ranked second and fourth among all databases as of this month, while MariaDB was 14th.
There are elements that could slow the expansion of open source relational databases. For example, Athena IT's Sherman said open source vendors and the communities that support the technologies could find it hard to match the pace of database innovation set by Oracle and Microsoft. But even if so, "that doesn't mean their databases can't be used for a lot of things," he conjectured. "There are a lot of small and midmarket customers and many large database applications where they will be more than enough."