The good folks at Cloudyn announced recently that they have added support for Google Compute Engine to their cloud monitoring and cost allocation service. As part of the rollout of that announcement, Cloudyn received some nice coverage over at GigaOM, specifically for a feature that compares workloads you currently have running on AWS and tells you what you would be paying for that same workload on GCE. In fact, they showed this really nice screenshot demonstrating the comparison:
While this is really useful information, is price enough to justify a move? What about performance, specifically the price-performance ratio? How would you go about choosing instance sizes on this new cloud? And if you were completely invested in AWS deployments to the tune of $7,000 month how would you go about moving to GCE?
You decided you want to move between clouds, now what do you do?
The Answer: CliQr’s Unmatched Portability
While it’s true we are partially funded by Google Ventures, CliQr and our CloudCenter product has been fundamentally about cloud portability since our founding. We believe that an organization wins when it has the ability to not just move, but manage workloads on any cloud they want, private or public. Application needs and cloud features are constantly changing and what was a great decision last year might not be so great right now, as the Cloudyn example demonstrates.
So why lock yourself in? Why not give yourself the freedom to choose? Then when the circumstances change, you can change with them.
How We Do It: Cloud-Agnostic Profiling
Our approach is based on creating a cloud-agnostic profile that describes the layers and requirements of your application. N-tier Java app? No problem. Rails? We’ve got you covered. Batch mode jobs or Hadoop processing? You can do that too.
We not only support a variety of common architectures, you can create a custom profile that meets your specific application needs. You can even automatically generate a CliQr profile from an AWS Cloud Formation template.
Let CliQr Embed Cloud Best Practices
The profile that describes your application infrastructure needs is used during the deployment process by our backend, CloudCenter Orchestrator, which has embedded in it knowledge of each cloud it runs on, including best practices. In fact, we have CloudCenter Orchestrator instances running on AWS, GCE, Windows Azure, Rackspace, and HP, and have them for OpenStack, CloudStack, and VMware too. That means, after you go through the one-time profiling process, you can move your application to any of those clouds (and more as we add them) in a single click.
Give Us a Try: Benchmark Your Application Today
Knowing when to change cloud providers and putting in the effort to actually change are two very different things. Cost of equivalent instance sizes is certainly an important factor into that decision, but generating benchmark data for your exact application so that you can size the instances more precisely is important as well. Because performance varies so much by cloud and by instance type within a specific cloud, the only way to know for sure what the best price-performance ratio is for any given task is to test it.
So, not only does CliQr CloudCenter make the migration easier and permanently make your application more portable, but it also includes a benchmarking feature that enables you to test your actual application with a JMeter script on multiple clouds with instance sizes you choose so that you can get a better determination of where you should run your workload and with what sizes. We’ll even generate a nice graph for you showing the deltas, not just in the performance, but in price as well so you can make an informed decision on which cloud and instance type is right for your situation.
And then when it is time to actually move, with your application already profiled, making the change to deploy production to a different cloud is a single click. Simple, portable, leaving you more informed and with options in the future. Try us out today and see for yourself!