Atlantis Raises the Software-defined Storage Bar

As Storage Switzerland covered in a recent briefing note “Solving the VDI Storage Contradiction“ Atlantis ILIO is a software solution that addresses the often conflicting requirements facing many companies with their VDI deployments: providing adequate performance for virtual desktops and doing it at a reasonable cost. The Atlantis ILIO software uses intelligent data reduction technologies to minimize the runtime ‘footprint’ of virtual desktops, enabling them to be run in server memory instead of on shared storage. This eliminates the performance issues of the storage network and the high costs associated with other acceleration methods.

Atlantis ILIO USX

Now Atlantis has taken this technology they developed for the VDI space and applied it to the virtual server environment. Atlantis ILIO USX is a software solution that runs in VMware and XenServer hosts, creating an abstraction layer that aggregates existing storage resources into shared pools. This capacity can come from internal disk drives, from shared storage on network-attached disk arrays, from server-side flash or even in-server memory.

These pools are aggregated across servers and network attached storage systems and dynamically combined into virtual volumes of storage at the VM, host or cluster level. Each volume can be created with specific characteristics to match the storage requirements of the workloads being supported, in terms of performance, capacity, availability, cost, etc. This means that higher performance applications will no longer need dedicated server/storage building blocks.

Software Defined

Instead of buying storage assets for specific applications, with characteristics tailored to those applications, ILIO USX allows users to create or define volumes from these pools that are ideal for specific workloads. It’s truly software defined storage. For example, an admin could define a performance tier out of server-side flash that’s installed in servers in a SQL cluster running critical applications and then set up a capacity tier on shared storage for backing that database environment up. Or a high performance application that doesn’t need a lot of capacity could be run completely in DRAM memory, or flash provided by the pool.

Abstracting storage resources from the hosts that consume those resources isn’t a new idea and typically hasn’t been applied to environments like this where storage performance, and latency, can be an issue. But the technology that makes this work is Atlantis’ ILIO optimization layer running on each virtualization host.

The Atlantis ILIO software intercepts I/O traffic at the hypervisor level (it installs as a virtual storage appliance on each host) and delivers I/O optimization for each workload, including in-line deduplication, compression and write coalescing, to dramatically reduce data that actually gets written to shared storage, by up to 98%. This ‘content-aware’ technology leverages the inherent duplication that exists in virtual machine images to service the vast majority of data requests from local storage. This enables the system to maximize the effective local capacity and performance, making the storage pooling concept work.


VMware’s vSAN is a similar concept, sharing resources and allocating them dynamically via software. But it’s not designed to aggregate storage outside of the server cluster, like to a network attached storage system. It also doesn’t do any optimization which is the secret sauce of Atlantis’ technology. By running this optimization process on each host server Atlantis can reduce the amount of physical data that actually hits the network between servers and storage resources. This can reduce network traffic and latency to storage systems, deliver more performance as well as extend the life of those resources.

However, Atlantis can still deliver its optimization benefits to vSAN, treating it as just another aggregation of performance and capacity resources, while at the same time adding shared storage to the resource pool. This could make vSAN more scalable and applicable for larger workloads than the current eight-node limit.

Storage Swiss Take

As we discussed in a recent webinar “The Truth About Software Defined Storage”, (actually our highest attended webinar to date), software defined storage is a term that’s become quite popular of late. This is probably due to its promise of increased flexibility, better performance and greater utilization of storage resources. But the term is losing some of its meaning because it’s being used to describe a wide range of products; products that bear little resemblance to what we consider the primary characteristic of this technology, abstracting physical storage from storage services. Atlantis’ ILIO USX is different.

By reducing the amount of data that has to be physically moved across the network, or even between servers in the same rack (although that’s still on the network), Atlantis ILIO USX enhances the software defined storage term instead of watering it down. These enhancements provide real benefits for virtual server environments and virtual administrators should take a hard look at this solution.

Atlantis is not a client of Storage Switzerland

Eric is an Analyst with Storage Switzerland and has over 25 years experience in high-technology industries. He’s held technical, management and marketing positions in the computer storage, instrumentation, digital imaging and test equipment fields. He has spent the past 15 years in the data storage field, with storage hardware manufacturers and as a national storage integrator, designing and implementing open systems storage solutions for companies in the Western United States.  Eric earned degrees in electrical/computer engineering from the University of Colorado and marketing from California State University, Humboldt.  He and his wife live in Colorado and have twins in college.

Tagged with: , , , , , ,
Posted in Briefing Note

Enter your email address to follow this blog and receive notifications of new posts by email.

Join 22,247 other followers

Blog Stats
%d bloggers like this: