At VMworld last week I had the pleasure to sit in EUC1494 – Plug and Scale “View Appliance”. This first half of the session the presenters talked about the Rapid Desktop Program and a in house VMware View Config Tool. The tool is for fast View installation from proof of concept to scalable production. The VMware View Config tool was code named Catalyst. Project Catalyst was also the same name used by VMware for a orchestration tool to create Multi-tenant VDI environment using VMware View and vCloud Director(also a great session). The VMware View Config tool had used the name first so I guess it wins! The View Config tool will add hardening practice guidelines and security integration in the future to it’s baseline installation.
The tool allows for a standard VMware View installation in 90 mins to give a fair baseline for all the participants in the Rapid Desktop Program. I think it’s fair mentioning that Rapid Desktop Program is a pass/fail program. It’s like getting your VCP, doesn’t matter if you get 300(bare minimum) or get a perfect score. The Cisco Integrated Services Router Generation 2 (ISR G2) has passed certification into the program and it can only handle 15-20 sessions. The Pivot3 vSTAC can handle 100 sessions per appliance. The 100 sessions are also based on a 1 vCPU number which I would say is not conducive to a good user experience. The Pivot3 can also scale up to 1000 sessions by adding more units but Nutanix can scale to 3,000 sessions and beyond. In the Rapid Desktop Program there is a big range of appliances based on overall IOPS, capacity and features. As long as you don’t have your blinders on, you’ll be fine. It’s great to see Vendors and Vars building solutions and not battling my array is faster than your array.
I walked away with two things from this session, the appliance model for View can in general can yield a 30% cost savings from reduced complexity and hardware. The other thought from the session was that you should try to use the appliance model for all your infrastructure. The appliance model can scale in known amounts, it’s easy to transition from Test -> Dev -> Production with repeatable results. In the cloud world it can be hard to predict capacity for future growth. It’s much easier to apply some clay to your sculpture instead of building a pyramid each time business requirements change.
Other Opinions on the Rapid Desktop Program: