Nick Toozs-Hobson

VmWorld: from a drawing on napkin to the best of breed unified platform

Blog Post created by Nick Toozs-Hobson on Oct 24, 2013

VmWorld in Barcelona last week was a seminal moment for me as Hitachi put it's promise on show to the 8,500 attendees. "Why would that be?" I hear you clamouring, well may be not clamouring, but perhaps someone may wonder why. In 2006 or 2007 I met a technical whizz kid who worked for Microsoft, in the MSN engineering team, on the HQ campus in Redmond near Seattle in the USA. His name is Rob Pike. Rob is a Canadian from Newfoundland who was responsible for building MSN Virtual Machine farms to weather the storms of demand generated over the internet for MSN's "properties". We were trying to convince him of the technical merits of our SAN-based products the Tagmastore (or USP as it was also known) and the Adapatable Modular Storage (AMS) on which to build these farms.

 

One day, when we were having lunch in The Claim Jumper restaurant in Redmond Town Center ( http://www.claimjumper.com/ VERY large portions) and Rob grabbed a napkin and drew out what he saw as the architecture for handling boot storms in a vuirtual world. He told us how Hitachi's technology was infinitely superior to all the other vendors he worked with (you know, all the usual suspects), and why our enterprise technology lent itself to building vast, rapid-to-deploy, VM farms at a far lower cost with higher consolidation than anyone else.

 

We subsequently arranged a series of meetings to introduce him to our Technical and Product guys, John Mansfield, Hu Yoshida, Sean Moser, Roberto Basilio and others to show them the architecture he'd drawn out on the napkin, I think I actually gave John the napkin!!. Anyway, Rob then worked with us both in Redmond and in Santa Clara to find ways of delivering hundres of VM's and rolling them out within minutes. That became his mission in life. About 2 years after we met Rob joined Hitachi and Project "Newfoundland" was born. We opened a lab in Bellevue WA (next to the Microsoft campus) and started to build what we know today as the Unified Compute Platform (UCP). So imagine how great it was to see our products on show at VmWorld delivering on the promise of that napkin from all those years ago.

 

Sadly, Rob has moved on to other things away from Hitachi, but his legacy is with us today in the form of the fastest to deploy, highest density, consolidated VM farm on the market. Not claims I make lightly, in conversing with our competitors I found out to roll out a VCE v-block takes a minimum of 15 calendar days for a small one, and over 45 for a large one, compare that with 3 days for a UCP Pro..... a v-Block is also limited to the single V-Max frame... ours is attached to the VSP soit can attach to multiple (heterogeneous) frames meaning it can scale and grow, but to do it in a foot print of 6+ cabinates, as opposed to 10+.

 

How right Rob was, thanks mate, eh!!

Outcomes