Astute Boosts VMware I/O Performance 1,500 Percent

VMware-based virtualization could get a big boost with Astute Network's ViSX G3 accelerator, which provided one early adopter with 1,500% improved read performance on critical applications and accelerated performance on its most demanding VMs, as well as delivered a noticeable increase in the responsiveness of its database applications.

August 15, 2011

3 Min Read
NetworkComputing logo in a gray background | NetworkComputing

VMware-based virtualization could get a big boost with Astute Network's ViSX G3 accelerator, which provided one early adopter with 1,500% improved read performance on critical applications and accelerated performance on its most demanding virtual machines, as well as delivered a noticeable increase in the responsiveness of its database applications. The customer, Visioneer, also reported that backups were completed faster, with less impact on other operations, and that productivity and efficiency were up, as well.

Company executives say this is a new category of solution, network performance flash, which makes virtualization affordable for smaller businesses. It allows customers to dial in the level of performance they need for specific applications and guarantees them the level of performance they're looking for.

Available in three models, starting at $29,000, the VMware-certified ViSX G3 is a combination of enterprise flash memory, the company's DataPump Engine processor technology, and custom hardware and software. Astute says the DataPump Engine processor delivers fully offloaded and accelerated network traffic (TCP/IP) and virtualized data store traffic (iSCSI), sustained flash performance and multilevel RAID protection--the sustained performance equivalent of 100s of enterprise disks.

Mike Karp, VP and principal analyst, Ptak Noel Associates, says this solution has the potential to be a great enabler for virtualization, whether that virtualization is inside the data center or in the cloud. "Its key strengths lie in several areas. First, it's the easiest way I know to improve performance in a virtualized environment. The process is essentially this: Plug it into the Ethernet, and walk away. Second, it's relatively agnostic, and out-of-the-box works with whatever network storage is available. Third, it provides increased IOPS performance across the network--not just to a single machine--and it does this at an aggressive price point."

Storage and virtualization solution providers are keenly aware of the IO performance issues impacting virtualization implementations and are actively developing solutions to off-set these challenges, but they have a ways to go yet, says Steve Brasen, managing research director, systems management, Enterprise Management Associates. "The new ViSX G3 platform could resolve this problem today and aid organizations struggling with 'VM stall,' boosting virtualization success rates and increasing the rate of adoption."

Karp's major concern is that this can be considered an "offload engine," which has had a bumpy ride in the marketplace during the last 10 years. "Historically, they have been expensive add-ons that have been added one-by-one to individual workstations or servers, and while they could scale, it was expensive to do so because you had to add new a one to each node where you wanted improved performance. Astute Networks will have to make sure people understand that while ViSX shares a common goal with the older offload engines, this is fundamentally different from that sort of technology: This goes out on the network, and can be shared/leveraged across whatever processes are being transacted there."

Brasen likes the fact that as an appliance-based platform, the ViSX G3 is a turn-key solution that is completely self-contained and can be easily and quickly deployed in new environments. "Although the initial release of the appliance is designed to support VMware implementations, future editions will provide more heterogeneous support, providing a single solution for supporting multiple virtualization deployment types. This heterogeneous support is critical as the majority of virtualization adapters (79%, according to EMA primary research) utilize more than one virtualization platform."

His caveat is that as an independent architecture, it adds an additional infrastructure component to the servers, storage and virtualization software platform that needs to be purchased and managed. "As is common in any virtualization implementation, organizations need to choose between performance and complexity. Though, in this case, I would argue that the performance boost far outweighs the increase management impacts."

See more on this topic by subscribing to Network Computing Pro Reports The Data Mastery Imperative (subscription required).

Read more about:

2011
SUBSCRIBE TO OUR NEWSLETTER
Stay informed! Sign up to get expert advice and insight delivered direct to your inbox

You May Also Like


More Insights