NuoDB 2.6, Part 2: Continuous Availability with AWS Active-Active
Released last week, NuoDB 2.6 features formal support for active-active deployment across Amazon Web Services Availability Zones. What does this mean for you and your applications? First, let’s talk about what active-active benefits can do for your deployments.
For mission critical applications that cannot lose data, the traditional high availability (HA) architecture consists of an active-standby database configuration. In this model, applications read and write data to a database instance, the so-called “active” database,with the hope that the “standby” database instance can take over the application workload if the active database fails. In the case of a failure event, there is typically a period of time between the failure and when the standby database can actually start processing the application workload. This gap in time before the standby database becomes active results in users without service until the database can be brought back online. In today’s online world, everyone expects always-on access to applications and uninterrupted service. So, even the smallest amount of downtime can result in your users ditching your application in droves.
The biggest attraction of active-active deployments is that they can eliminate downtime for both unplanned and planned outages. Since both database instances are active, service continues uninterrupted if one database goes down. Active-active deployments also allow you to make use of all of your infrastructure versus having it idle as insurance in case of a failure. But achieving these benefits isn’t always straightforward. Traditional databases only provide active-standby functionality, requiring you to purchase expensive add-ons to enable active-active capabilities. NoSQL solutions offer high availability but don’t offer true consistency or other ACID-guarantees, forcing you to compensate by writing specialized data management functionality into the application.
With NuoDB, the benefits of always-active, always-available data are a natural result of NuoDB’s peer-to-peer, distributed, two-layer architectural design. We provide active-active benefits, without the complexity of a traditional add-on approach. How do we do this? NuoDB acts as a single, logical SQL database, no matter if it’s deployed in a single datacenter or across multiple data centers. Since it's a single database, applications don’t need to be aware of, or have to grapple with typical active-active complexities such as coordination of independent active databases in different locations, conflict management, scale-out, or the other typical active-active complexities. This frees up developers to focus on the important stuff - coding application features and functionality that really matter to the business.
Under the hood, NuoDB can be deployed within a data center or across multiple data centers, while maintaining ACID guarantees and serving up data at in-memory speeds. With this set-up, if one data center suffers a disaster, the other can seamlessly take over the workload with no downtime.
Here’s a diagram on how NuoDB 2.6 can be deployed in AWS as an active-active set-up:
NuoDB is a single, logical database that can be deployed across multiple AWS availability zones, providing active-active benefits to applications in each zone.
Want to learn more about active-standby, active-active, and NuoDB’s approach? Look for a deeper dive series on the subject coming soon! We’ll go over the traditional approaches of implementing active-standby and active-active environments along with the benefits, drawbacks, and considerations of each approach. And we’ll get deeper into the details of NuoDB’s approach to active-active deployment, how it’s unique in the database industry and what you need to consider with NuoDB and this type of setup.
In the meantime, check out our white paper on Continuous Availability in NuoDB.
This is Part 2 in a multi-part series getting in-depth with NuoDB 2.6! Check out Part 1, Introduction NuoDB 2.6, to learn everything we released in NuoDB 2.6 and look out for additional blogs about Table Partitioning and Storage Groups (in two weeks!) and SQL Enhancements coming out shortly after.