Google and Why the New Standard for Modern Applications is a Non-Relational Database Deployed in the Cloud #CloudWF
Guest Blog with MongoDB
Google and Why the New Standard for Modern Applications is a Non-Relational Database Deployed in the Cloud
Author: Kelly Stirman, VP of strategy at MongoDB
It’s positively raining cloud stories. Sorry. Cloud puns are so over…cast. Regardless, recent months have seen some interesting developments in the high stakes game for control of the foundational layer of your application stack i.e. what database you use and where it’s deployed. In early May Google released Cloud BigTable as a managed NoSQL database. Two weeks later Gartner released its Magic Quadrant for Cloud Infrastructure as a Service (Cloud IaaS) report.
While unrelated, the two announcements both shine some light on our path to a new, cloud-rich future. While the aspiring cloud giant Google gave further validation, if it were needed, that NoSQL databases deployed in the cloud are the new standard for modern applications.
You see, the workload from modern applications is quite different from what it’s been in the past. Building your own data centre and installing a relational database was fine when you could predict the size, speed and type of data. Applications in 2015 are a different breed. The growth of social, mobile and sensor data has dramatically altered the way we approach development. Developers can’t tell in advance what any of this will look like in the final production version of their application, let alone future iterations.
Many organisations are already overcoming this by deploying non-relational databases on commodity hardware in the cloud. This approach lets companies gear up for massive scale and gives them enough flexibility to incorporate new data types that will support business processes and provide operational insight.
Google’s BigTable and Gartner’s Magic Quadrant
Google’s announcement highlighted two things. One: the big infrastructure players are looking to diversify and find new ways to wring revenue from the big data stack. Two: BigTable’s release illustrated that all major data innovation is happening away from relational data models. Relational databases aren’t going anywhere fast, but they are challenged by the requirements of modern applications. In particular the trickiest of the three Vs of big data – variety of data types. BigTable is yet another database-as-a-service offering that is designed to be deployed on the vendor’s own cloud infrastructure, see also Amazon and Microsoft.
From one cloud provider’s announcement, we now look at a broader view of the industry from Gartner. This is from the introduction to the Magic Quadrant for Cloud Infrastructure as a Service, Worldwide report:
The market for cloud IaaS is in a state of upheaval, as many service providers are shifting their strategies after failing to gain enough market traction. Customers must exercise caution when choosing providers.
The report went on to explain that ‘all the providers evaluated are believed to be financially stable, with business plans that are adequately funded. However, many of the providers are undergoing significant re-evaluation of their cloud IaaS businesses’. In other words, some vendors may not be in it for the long haul.
What it means for you
As well as diversification into database services, the cloud competition is also sparking a healthy price war. Just a few days after the Magic Quadrant was released Google announced it was slashing prices by as much as 30%. Microsoft and Amazon are also fond of aggressive pricing as they try to eat as much market share as possible.
Which brings us back to Google’s launch of NoSQL database-as-a-service BigTable. The release came on the back of Microsoft’s recent Azure DocumentDB announcement and, of course Amazon’s own DynamoDB offering. As the competition for cloud infrastructure drives margins down, the big players are looking up the stack to drive revenue and it’s clear NoSQL technology is one of the most attractive areas.
Though it’s worth pointing out that these as-a-service database offerings generally come with a very narrow set of features. For example Cloud BigTable is a wide column store with a simple key-value query model. Like some other NoSQL databases, it is limited by:
- A complex data model which presents a steep learning curve to developers, slowing the rate of new application development
- Lack of features such as an expressive query language (key-value only), integrated text search, native secondary indexes, aggregations and more. Collectively, these enable organisations to build more functional applications faster
Ultimately the cloud providers can relieve users of some of the overhead of running a database but they still will have to deal with the complexity of mastering data models and working around key-value query limitations.
Out of the chaos it’s becoming clear that a non-relational database hosted in the cloud, is going to be the predominant way modern companies deploy applications. Each customer will have varying demands of control. Some will want everything ‘as-a-service’, others will want full control over how and where their database runs and security on each layer of the stack. In the modern world of cloud-ready, non-relational databases, you have more choice than ever. That choice can also bring a risk of vendor lock-in, if you select an offering that is tied to one specific platform, no matter how ‘web-scale’ that platform claims to be.
 Gartner, Magic Quadrant for Cloud Infrastructure as a Service, Worldwide, Lydia Leong et al, May 18, 2015
MongoDB will be exhibiting at the Cloud World Forum taking place on the 24th & 25th June 2015.
Kelly Stirman is Vice President of Strategy at MongoDB, speaking at the Cloud World Forum on the 25th June at 10.35 in Theatre A: Keynote – Building Business in the Cloud on ‘Escaping Cloud Cuckoo Land: 5 Tips for Making Success a Reality in the Cloud.’