The world shall be mine!
I had another moment of epiphany yesterday about how UCS is less about infrastructure and more about delivering value to the business. How can “just another bit of tin” be any different? Read on.
First I read multiple articles and had multiple discussion about VMware’s acquisition of SpringSource (what the hell does that have to do with UCS? Everything!). For us old time, ex-VMwarians this acquisition makes a lot of sense because we were already focusing beyond the infrastructure teams and on the executives, business and applications folks. That’s why Propero and Dunes were bought. VMware is focusing on the return on investment because IT should be about delivering value not just eliminating cost.
Then I read a fantastic article by Steve Duplessie at Enterprise Strategy Group (thanks to Ian the Grumpy Scottish Pessemist Storage Guy for tweeting it) on Scarcity Imbalances and what really matters is not “cost per CPU cycles” or “GB of storage” (because they are abundant), but time and productivity (because they are scarce). To paraphrase Steve:
Anything you can do to save me time (and money) is a no-brainer. Make IT simple, stupid.
UCS has a number of unique selling points that enable this goal to be achieved:
- Simpler. Making the data center simpler with commodity, standards-based architecture means it is easier and faster to deploy compute, network and storage for the applications folks. UCS is new and different, which people equate with complicated: but it really isn’t, it’s just a big host, really.
- Dynamic. UCS was designed for virtualization and automation. Imagine running many varied workloads on vSphere on UCS, mixed in with non-virtualized applications like grid – and this can change by the hour, via a portal, with no complex IT interaction.
- Self-service. Everything is automated when you add a self-service portal to the front end – you don’t get a self-service portal with UCS, but the system has been written to be open and remote controlled through the XML API. This means Cisco lets YOU choose YOUR portal instead of forcing you to buy ours. Win-Win.
- High performance operations. Streamlined, pro-active operations now becomes a priority. Unplanned, hair on fire, problem-driven operations does not add value and is a waste of money. An example is capacity management: when IT works to ensure that the application team has all the capacity they need, then they are adding value and being proactive.
- 3D scaling. There is a gap between IT and Applications, and it varies from org to org. This is characterised by IT not being able to predict the capacity demands of applications teams. With UCS it is easy to scale out (more blades) and scale up (more memory) and scale deep (virtualization).
I no longer look at UCS as an infrastructure component: I look at UCS as a value-creating, application-enabling, knowledge-worker helping system that a CIO can brag about on the golf course.
If this sounds like marketing BS to you, then I’m guessing you work in IT, probably in one of the silos, and you’re probably operations focused. Am I right? This is real, customers are building these systems today and this is the route to the cloud. You don’t have to believe me, but I WILL be saying “I told you so!” and dancing a merry jig at some point soon 
UCS can run desktops, it can run Oracle, it can run vSphere, it can run Grid: and it can run them all together, or scheduled at different times – you make the rules.
Related posts:
- VMware + Ubuntu + Ruby + REST + XML + Cisco UCS API
- Cisco 6100XP aint Switches like VMware ESX aint Linux
- Cisco Solutions for a VMware View 4.0 Environment Design Guide
[...] UCS systems being delivered to a client of mine last week, I found Steve Chamber’s post about The stars are aligning: VMware, Cisco and ESG rather timely and right on target with regards to the higher level value UCS delivers to the [...]
Steve, great post and definately in agreement with you. I do wonder though, especially with Cisco and EMC pushing their own pieces of kit hard to customers what the difference really is between buying a vBlock from VCE and buying a V+C+E solution? I’m not talking about the obvious selling point differences ( I flat out agree with all of those!) but wonder a; what the actual difference is from a BOM point of view and b; what the price difference is when you compare a vBlock stack to the same kit but coming as a Cisco, EMC & VMware solution…? I would take a vBlock hands down because I have been on the pointy end of trying to manage desperate compute, storage and network IaaS stacks but would like to hear what you say to customers who ask the same question I have…?
you have to look at the whole process from start to finish and find benefits along that path. For example, if you tell me some specific inputs (e.g. VMware capacity planner data) then I can create a Vblock BOM that you can order in < 5mins. That's ~198 line items that will be shipped NOT in 198+ boxes, but on ~3 pallets. Then there's the solution-specific bundles you can buy (e.g. Cisco Unified Comms blade packs, already configured to meet Cisco's Voice Technology Group (VTG) standards), or how about the "one throat to choke" seemless support, or how we can ship around the world to any datacenter? I could keep going on.
I've built "franken blocks" before I joined VCE and I know how hard V+C+E is.