Architecting for the Cloud Series – Lessons from the Experts
In March, NuoDB held the first in our 2015 breakfast seminar series, “Architecting for the Cloud” in Cambridge, MA. Thanks to our database enthusiasts and experts, the event was a great success. Read on for the highlights or watch the video recording of the event below!
Our guest speaker, John Treadway, SVP of Cloud Technology Partners, started the discussion on cloud database architecture by walking through the pre-cloud applications era and its database architecture, and how the evolved information technology requires new, more agile database architecture.
In his comparison of the old and new ways, John brought up several points that are quite noteworthy:
Microservices are powerful, but use them wisely
For those of you who aren’t familiar with the term yet, microservices, as defined by Martin Flower, describes “a particular way of designing software applications as suites of independently deployable services.” John noted that while microservices are “very powerful”, tracking all the services is complex work. When done the right way, the extra scaling ability and agility it offers will allow developers to update services quickly without taking down the entire application.
“It’s not how big you can get; it’s how small you can get”
“How quickly can you get small when the demand goes away?” Through this thoughtful question, John brought up a quite inspiring point: despite the trend to “scale up” and “scale out,” enterprises really need the ability to scale-in and scale-down to avoid overprovisioning and ensure database capacity adjusts to needs. John states that ultimately it’s about garbage collection in your VM farm. How do you know which instances are not running anymore? How do you make sure that when you take down a server or when you request information from that server, that you actually get what you want?
Disaster recovery vs. continuous availability
The concept of disaster recovery should sound quite familiar to all of us; it is based on the occurrence of failures. However, in the modern world when there are so many transactions going on at the same time, companies cannot afford to lose the ability to keep up with those transactions. Despite continued efforts and assurances by high-profile cloud service providers, there is no shortage of well-publicized cloud outages, so the need for continuous operations had grown more and more apparent to enterprises.
Our CTO Seth Proctor added to the importance of continuous availability, providing his thoughts on what resiliency means to businesses. “Why is the dial tone always there? Because the telecom industry designs for zero down-time,” said Seth. Making the point that cloud and distributed computing are more than just JSON queries and lines of codes, Seth asserted that it’s all about the architecture.
Even after the event, many attendees came up to Seth to ask about our architecture and the continuous availability that it delivers. To learn more the topic, download the Continuous Availability Whitepaper.
Thank you to those who made it an awesome event, please stay tuned for more updates!