The World is Waking Up to the Value of a Database Abstraction Layer
When we set out to build ScaleArc almost five years ago, the entire scalable database landscape was shaky and fragmented, with various divergent approaches being taken by different innovators to address the scalability challenges facing anyone who wanted to build applications that could serve millions of users on the web.
There was a lot of noise around NoSQL, next-gen databases (VoltDB, Clustrix), and big data (Hadoop). We put our heads down, and spent nearly two years in deep R&D. Whenever someone would ask what we were working on, the answer was “We are building a database scalability technology.” They’d assume “you’re like X,” to which I always responded with “We’re nothing like that,” which inevitably confused people. Let me explain.
We had decided from the beginning that we didn’t want to build a “new database”. We believe it’s the apps that drive the database, and not the other way around. Applications have to be re-written to support new databases – the top 20 MySQL applications from 2008 are nearly the exact same set today, despite the availability (and hype) of MongoDB.We wanted to build a tech that could be used today – for the apps that run your business and generate revenue today.
Second, and this was controversial, even with our own investors, we didn’t want to build our tech on top of an existing open source engine that might be faster, but it would lock you into one database. We wanted to build something that could be of value to MySQL, Microsoft SQL server, Oracle customers, and some day, even more databases.
Third, we said that you could bring the performance that NoSQL provides to the SQL stack, and make it happen without a single line of code change in apps. Plus we wanted to bring transparency and visibility to the database stack, which has traditionally needed “Black Magic” from DBAs to maintain.
All of these restrictions meant that our tech had to work with existing databases and apps, yet still provide the same kind of performance, scale, and availability that would have otherwise needed a full-scale rip and replace of the database stack and the app.
That was five years ago. The world is different now. In 2010, we filed our first set of patents for this tech. Then in 2011, other projects started showing up which were trying to do similar things, though for a specific database or another. 2012 was when this new market category we now call “Database Traffic Management” became clearer, and 2013 was when customer acceptance and use cases started bubbling up to the top. Early in 2014, we see the rest of the world has woken up to the value of the database abstraction layer, with two different MySQL open source projects (MaxScale and ProxySQL) launching to support a subset of the use cases that ScaleArc has supported since 2011. Additionally, commercial load balancing vendors such as F5, Citrix/NetScaler, and A10 are launching basic database load balancing.
Having seen this need early, and developing our tech very aggressively to meet ambitious goals, ScaleArc is ahead of the pack in terms of capabilities and proven technology as the only solution that provides load balancing, auto-failover, dynamic caching and SQL query routing and sharding, and deep SQL analytics across MySQL, Microsoft SQL server, and Oracle.
The pace of innovation and customer adoption is only speeding up!
comments powered by Disqus