Corporate Blog

Smarter Tech, Better Analytics with Cloud-Based DataMine

This blog post originally appeared on turn.com. Turn was acquired by Amobee in April 2017.

As anyone who has gotten into a schoolyard argument knows, there’s nothing greater than infinity. Pull out the biggest number you can imagine and you can always top it with the one that no one can imagine.

That, in a nutshell, is the appeal of cloud computing. Though not actually infinite, the capabilities of cloud computing approach infinity in ways that local data centers can’t.

This logic underscores Turn’s latest decision to move DataMine to the cloud instead of hosting it locally. For our customers, this is great news. Because there is more computing capacity that scales quickly, queries will be faster; insights and analysis will be quicker; disaster recovery will be standard; and pricing will be more flexible.

Here’s a closer look at the benefits:

  • Flexibility: For customers, this will be one of the most noticeable changes. Previously, a customer could commit to a fixed set of nodes, but some capacity could go wasted when no queries were running. On the cloud, capacity is used only when a query actually runs. Additionally, the cloud allows for more elasticity, meaning you can commit to a certain number of CPU hours per month with the option to pay as you go if you need more. And because we can offer more nodes, customers can launch multiple queries at a time now without waiting for them to queue.
  • Speed: Queries will easily be 2-4X faster now. You might ask, “Why not 10X?” The answer is, creating a 10x larger cluster provides diminishing returns as the cluster size grows, while the cost scales linearly. But theoretically we could create a huge enough cluster to run any query within an hour. Deployments will also be faster because there’s no additional capital expense and no need to order new hardware. Instead, we can invest in features and capabilities to help clients and yield insights much faster.
  • New UI: A new UI will let marketers monitor and manage capacity. The UI allows more control over usage of capacity and the ability to order more as needed. The fixed capacity framework wasn’t as easily accessible to smaller brands, so now access is opened up to brands of different sizes.  
  • Disaster Recovery: Because we are using the Google Cloud Platform, all of our data will automatically be copied in different data centers at no extra charge. That means by default our customers get disaster recovery when we move into the cloud. From a client’s perspective, the data is always available. If something goes wrong with one data center, we can always use the other one.

Implementation

Over the long term, moving to the cloud means a big boost in performance and more power to experiment. Best of all, it’s a seamless transition. Clients won’t notice any changes on their end except that they’re able to do more. To infinity and beyond!