The race is on: EMC has announced a Vblock for VMware View on the same day that news broke of NetApp, Cisco and VMware's FlexPod for VMware. Now we'll see which Cisco and VMware partner, EMC or NetApp, can develop integrated specific system application and hardware stacks faster. Both EMC and NetApp are partnering Cisco and EMC …
Cisco the floozy!
Anyone buying into either of these stacks at the moment is taking a huge risk - first of all you are assuming that both EMC and NetApp will continue as entities in their own right. What happens, for example, if IBM buy NetApp? Also, maintaining a relationship and a bit on the side is tiring business ... can Cisco be giving full attention to both? Can both relationships stand the test of time?? And how integrated can 3 separate companies really be??
So what do I get by buying a Vblock, VCE stack, VCN stack or whatever they are called these days! Over buying the same components seperately? What functionality do I get that is different because the overall solution is greater than the sum of its parts? The answer is bugger all! All they have managed to do is let a customer buy a pile of bits with one part number something that SI's and proper vendors have been doing for years.
I hate it that they throw the term "cloud" around as if they actually know what is required to build one. The trick to making a cloud work is all in the software stack (orchestration, portals, catalogues) yet at no point have any of the vendors/coalitions mentioned in this article actually addressed this point. Hardware actually is a secondary consideration, as a pile of bits from three different vendors is still a pile of bits from three different vendors even if you have bought it from one supplier. What you really need is some form of over arching software stack that ties it all together (and does it out of the box, there should be no customer specific integration work required) and thats what HP and IBM's solutions offer. The solutions should be greater than the sum of its parts. Oracle do not have a cloud offering BTW!
There are three features that are a MUST in a cloud environment, inbuilt orchestration, inbuilt service catalogues (that should take into account apps) and service portals. Show me were this is in any of the products mentioned in this article?! All you really need in the hardware is the inbuilt integration points to let the sotware hook in and drive it.
All Vblocks and FlexPods are, are static bundles of hardware with none of the actual components required (pre integrated and out of the box) to do anything remotely cloudy!
"Every supplier involved in the Vblock and FlexPod efforts will be privately pleased that all the others are there at some level. This demonstrates a degree of openness that is missing from, for example, the HP and Oracle integrated stack offerings."
No it doesnt!! Openess means the ability to rip out one vendor and replace with another WITHOUT changing the architecture or removing functionality, Could I for example use HP servers instead of Cisco in a Vblock? Could I use IBM storage instead of EMC, could I take a VCE environment and replace the E with an N (EMC with NetApps for those not paying attention) without affecting the rest of the stack or having to rework something. The answer is no, these systems are far more closed than any solution out there. Let me give you an example can the customer choose which hypervisor to use? A hypervisor is just a tool for resource pooling so why do I have to use vmware when Hyper V is probably good enough? A decent stack should allow the customer some leeway in individual components. There is no true open stack out there but some are more open than others.
Do some research
Adam - I suggest you go and take a look at EMC's UIM and you might find the answer to your question. Just because this article doesn't mention it doesn't mean that the orchestrtion piece doesn't exsist.
Rubbish - YES
Again and again , the IT Companies are Just developing new ways to market the same Old technology , no real Innovation here . Most of them are just Marketing companies rather than Engineering company .
With too many complex marketing , pricing wars, the customers are unable to focus on the real IT issues , and Buys this Dump Boxes. I recently come across a Service provider Bought a 10M+ offering from one of this consortium , only to find 32 CPU's inside the Box . (CPU in a Cloud of storage ) .
Even the rare few Innovations , we Unfortunately fail to acknowledge. I.e ZFS from SUN changed the way Oracle could improve database Performance , the JUNOS from Juniper changed the way you could address security .
Why cant we replace TCP / IP with a new protocol , Why not a properly re-engineered Ethernet like Medium for Network & storage
IT has become a Industry driven by pure marketing gimmicks.