Scale-Out Architecture, Quality of Service, and Polyglot Persistence – a story of databases
Today, April 8 2014, we announced a series of best practices for deployment of Oracle & Microsoft SQL Server. This is an extension of the work that SolidFire has previously performed with MongoDB. The press release – which was entitled SolidFire Widens Reach In Enterprise Database Market with Oracle and Microsoft SQL Server Solutions, Oracle Gold Partnership – includes a quote from Jeff Wright (Database Application Engineer) that is quite compelling. A section of it, in particular, resonates strongly with me.
“SolidFire guarantees performance from tens to hundreds of databases within a shared infrastructure, facilitating significant storage consolidation, improving query response times and shrinking backup windows.” – Jeff Wright, Database Application Engineer, SolidFire
In 2011, Martin Fowler coined a term – Polyglot Persistence – that radically impacted my approach to the consideration of database design and usage. The notion is fairly simple, that any enterprise will be using a variety of different data storage technologies for different kinds of data.
Enterprise-IT, particularly in the area of data persistence, has undergone a massive transformation in thinking. I may deploy an OLTP system when consideration of consistency and transactional control is a business requirement, but utilize a “NoSQL” system when my data model is inherently non-relational, can be easily denormalized, or development patterns dictate. In fact, many enterprises utilize different persistence solutions in the context of the same application.
Early in my career, and during all of my database design & implementation courses, there was an inherent assumption that a “database” was not just a software solution but a monolothic set of infrastructure components including, in many cases, compute, network, and storage. Polyglot persistence that maintains the inherent assumption of monolithic infrastructure creates almost as many problems as it solves.
This is the beauty of the SolidFire storage system. Shared, scale-out storage infrastructure enables your model of polyglot persistence without investing in one-off, per persistence model, infrastructure solutions. Combining this with SolidFire’s granular, guaranteed performance Quality of Service controls ensures that this shared infrastructure is composable and tunable for the business, application, and development requirements faced by the enterprise IT department.
It is often no longer wise to invest, heavily, in a single database solution. Rather the forward looking enterprise – those designing and implementing their Next Generation Data Center – utilize database solutions that are tailored to access patterns, data models, and development preferences. The best solution for this polyglot persistence model, with shared storage infrastructure, is SolidFire.
-Tyler Hannan, Sr. Marketing Strategist