After finishing my share of tapas and fantastic fried octopus with potatoes, it’s time to reflect on the VMworlds of 2014, in particular the discussions I had with customers, partners and peers during meetings and conversations in the Solutions Exchange.
Before I start, let me tell you a little about my background. I’m a Global Solutions Architect with a past in IBM GTS. I have performed many data center consolidation projects. The first project that relied on VMware virtualization was back in 1999, with half the servers virtualized and 3-5 vms per ESX server –and it’s exciting how data center technology has evolved over the last 15 years. My main focus today is data center networking for both storage and application networks. I help develop and validate best practices for datacenter network design and data center interconnect networks for disaster recovery / business continuance (DR/BC).
VMworld in the US had its fair share of both product and visionary announcements, but the smaller European twin was more focused on solutions highlighting new products and announced enhancements. In Europe, my conversations were more about the best way to apply these to current data center designs and how the direction of infrastructure development will change with these new products and capabilities.Although the solution discussions in the US and Europe were very similar, in the US the scale is generally larger but the pace of adoption for the latest features and enhancements is a bit slower than in Europe, and even Australia. ”How Big” is dominantly a key requirement for US-based companies, though I also had a couple of discussions about using OpenStack. But in Europe, not a single customer talked to me about OpenStack. Perhaps the Europe show had more SMB companies than the US show so they deploy VMware suites of products and take more of an “all in” approach with a keen focus on achieving cost efficiency and differentiating capabilities in the market. This means European companies use rapid adoption of new features with full software suites to create their competitive advantage. I also got the impression that more folks in Europe perform tasks every day that span across technical domains than is the case for their US peers. Or you could say they have more of a cloud-architect/specialist job while their US peers operate mainly within a single technical domain.
Both in San Francisco, and Barcelona, I had multiple design discussions about how to connect two or even three data center storage and IP networks together. The main objectives were providing DR/BC capability and uniform deployment of applications across two or three DCs. This is often called Active-Active data centers, or with three, AAA data centers. For each customer, the motivation was uniform deployment of applications across two or three data centers with ‘built-in’ work load and data mobility between the different sites. In some case they wanted live vMotion but more common they wanted a guarantee that the application can be run in any data center and automatically included in DR/BC without requiring validation every time a new application is deployed regardless of its primary location.
And they all wanted high utilization of resources with utmost flexibility to utilize spare capacity regardless of which data center has it. In many of the design discussions a lot of questions were asked about how to take advantage of new features such as VSAN, NSX and EVO Rail in AA-DC (or AAA-DC) designs –since each of these features are not currently supported to be stretched across data centers.
I had the pleasure to work with Mohan Potheri (Sr. Solutions Architect, VMware) & Nanjunda Somayaji (Sr . Technical Alliance Manager, Brocade) on how to use a combination of in particular VSAN and NSX to enable mobility and DR from private DC to the cloud for Business critical applications for the following breakout session