Last summer, our senior management visited all of our most important customers around the world. We logged tens of thousands of airline miles in a dozen or more cities on two continents to hear what they had to say. The one thing our users kept telling us was that DevOps was still too hard. They didn’t want to think about recovering down nodes, shards, replicas, RAM, storage, and a host of other variables. That’s why today we’re introducing  Supergiant.

Supergiant is our new deployment and management technology that automates many of the most frustrating Elasticsearch DevOps tasks, achieves auto-scaling, and does all of this at a lower price point.

Why are we making these changes?

Before we go into Supergiant’s main features, we first want to go into some background that explains why we’re making these changes.

Since our inception, Qbox has been seeking to provide the best, most performant Elasticsearch in the cloud with the most accessible support, and at affordable prices. These offerings powered us to thousands of users, managing thousands of nodes and clusters. But there were still way too many variables and use cases to make Elasticsearch DevOps invisible.

Keeping our user’s needs in mind, we embarked on a journey to improve our cloud infrastructure, to make it more resilient and fault tolerant. Our main purpose was to allow our users to focus on building their apps instead of fretting over DevOps.

We wanted to do something dramatic: both lower prices and provide more computing power, but we couldn’t get there by using our old deployment model. So we reimagined our infrastructure from the bottom-up, started from scratch, and the result was our Supergiant infrastructure, a fundamental change to Elasticsearch in the cloud.

Supergiant goals

We started out with 4 clear goals:

  • All clusters should recover from failure automatically
  • Customers should be able to resize clusters seamlessly
  • Hosted Elasticsearch should be more affordable
  • Customers should be able to dedicate exactly the resources needed

And after months of dedicated engineering effort, we created Supergiant.

Supergiant features

With Supergiant, you can now expect automated disaster recovery, seamless resizing, and precise provisioning controls at a highly-competitive price.

Automated Disaster Recovery

In case of catastrophic server failure, Supergiant guards your data and relaunches new clusters, so computing power is continually assured with automated disaster recovery without manual intervention.

We engineered Supergiant to:

  1. Sense when a cluster is about to go down
  2. Re-provision computing power in its place
  3. Attach itself to the same data source in under 60 seconds.

Even on zero-replica clusters, no manual intervention is required to recover from a disaster, and on replicated clusters, zero-downtime performance is assured.

Automation Yields Seamless Scaling

Resize and replicate any cluster from your dashboard without migrations, switching endpoints, or additional manual intervention. Simply select the size that meets your changing performance needs.

No more switching endpoints when you’re ready to scale up. Even zero-replica development clusters can quickly add replicas and be resized to handle the heaviest of production query rates on demand.

More Affordable at Any Scale

Right-size your cluster to the correct number of CPUs and RAM independently, and never pay for extra deadweight infrastructure again. You can dial resources in to match the exact profile you need. Lucene shards are handled automatically for performance, hard drive space is infinite, and disaster recovery is automatic.

Approaching it this way made for better resource allocation on smaller clusters and even better efficiency at scale.

Check out our pricing page for more details.

Plus, Qbox Premium Support and Features

Additional to all of Supergiant’s features, you still get Qbox’s free 24/7 premium support, maintenance and diagnostics provided by trained Elasticsearch experts. Every person you interact with at Qbox is trained for Elasticsearch and nothing else. This just simply is not the case if you are self-hosting or using a thrown-together solution from one of the public cloud providers.

So if you haven’t started yet, get started today!

Contributors: Mark Brandon, Qbox CEO | Ben Hundley, Qbox CTO | Brian Sage, Qbox Product Manager | Andrea Trefftz, contributing editor