TidalScale’s WaveRunner – the point-and-click control panel that makes creating a right-sized Software-Defined Server fast, flexible and easy – isn’t just about creating one or more virtual servers from multiple commodity systems. It also puts you in control of all the software-defined pieces in the data center. So in addition to cores and memory, WaveRunner allows you to monitor and manage storage and networks. You simply pick the software-defined resources you need and plug them together.
Like Amazon’s EC2, only better
Customers of Amazon’s Elastic Compute Cloud (EC2), part of Amazon Web Services (AWS), are already familiar with this concept. EC2 gives users a way to pick a server size (essentially a virtualized machine running on a physical server), along with software-defined storage and a virtual private network devoted to your instance.
We love this concept, but we don’t love its limitations. And we expect we’re not alone. Because while EC2 is useful for leasing and operating a relatively small software-defined data center (SDDC), it won’t help users who need systems with a lot of memory or cores. That’s a real problem for the kind of use cases that rely on in-memory computing – applications like intensive simulations, bioinformatics, financial analysis, data sciences and business intelligence. As datasets continue to grow in volume and variety, and as models grow more complex, the standard server configurations available from EC2 just won’t do the job. Nearly all the instances available on EC2 are equipped with just 256GB of memory and fewer than 40 cores (see graph). For large in-memory computing tasks, EC2 offers just one system with more than 1TB of memory. And good luck getting your hands on that.
The essential problem that EC2 faces is the same that all data centers face; it’s a simple matter of scalability. EC2 offers access to single servers (or more accurately, virtual machines running on single servers). If your memory or core needs eclipse what’s available on a single server, you’re out of luck.
With TidalScale WaveRunner, built on TidalScale’s powerful HyperKernel platform, you have access to a different approach. Instead of running a virtual machine on a single server, which limits you to the resources available on that one system, WaveRunner allows you to configure all the resources of multiple commodity systems into a single server that perfectly suits your needs. With just a few clicks, you can choose the resources you want to combine from your datacenter, or from private and public clouds, and create a right-sized server that has all the memory, cores, storage and network that you need.
Scaling beyond a single system
TidalScale virtualizes all these resources and presents them as part of a single system, or as many as you like. So you’re not limited to the memory available in just one box (even EC2’s nearly 2TB system, if you can lease time on it). You can combine the memory of 20, 30, even 40 or more servers. If each of those servers is equipped with 1.5TB of memory, you’re looking at a Software-Defined Server with roughly 60TB of memory.
That’s an extreme example, but with WaveRunner as your control panel to right-sizing you have countless options for configuring all the software-defined resources that you need. You can do it on the fly, and then add or subtract resources as needs change. And it requires no changes whatsoever to your operating system or application software.
A better SDDC on demand
That’s the real advantage to TidalScale WaveRunner and its HyperKernel platform: flexibility. There’s no worrying about whether or not your on-premise datacenter or cloud provider has a large enough system, or enough storage, or the right network, to handle your workload. TidalScale makes it all part of your very own SDDC – one you can configure and control entirely on demand.
And with its support for several popular network switches, WaveRunner even saves you from having to become an expert on network operating systems from Mellanox Technologies, Juniper Networks, Cumulus Networks and Ubiquity. Simply select a supported switch from the WaveRunner control panel and automatically create a virtual LAN (vLAN). Now all your software-defined resources, including server and storage, are running on a vLAN that you built in a few clicks with virtually no networking expertise required.
We think Amazon’s EC2 is a fantastic offering for two reasons. First, it proves the need for users of all kinds to configure and operate their own SDDC. And second, it shows that there’s room for improvement. With TidalScale’s WaveRunner and HyperKernel, you can take the EC2 concept and unleash it in your own data center, using your own existing IT assets, and take it places even Amazon can’t go.
Watch this video to learn more about TidalScale WaveRunner.