Monday, 23 May 2011

Vertical stacks - and what I expect to hear.....

I am having a few “challenges” getting vendors to understand what would motivate me to purchase their own “vertical stack”…. Oracle, IBM, HP, VCE, Dell and others all have them – and yes, this is aimed at all of you!

When you come knocking at the door selling your wares firstly understand what you are selling and why I would be interested… Understand how the enterprise works and what integration needs to be completed.

What I expect to hear:

·         Yes all the interop testing is managed on my behalf

·         Yes all of the reporting / dial home etc is included

·         Capacity planning tools are included

·         Management is simple and done from one place. All aspects can be handled in one place

o   Hardware configs & management

o   Compute allocation

o   Network configuration

o   Storage configuration

o   Setup of storage replicas for both sync and async – to disaster recovery sites

o   Hypervisor

o   Virtual machine instances

·         Yes it is capable of production  / dr and we have standard patterns for this

·         Yes I can back the thing up

·         Yes I understand how I can build this into ANYONE’s infrastructure

·         All of the standard patterns we have mapped have full ISV support (and reciprocal support agreements are in place)

·         Standard patterns describing all use cases are available (anything out of these use cases is NOT a valid use-case)

·         Documented pre-req’s for my infrastructure – on what is expected to make this work


What I expect NOT to hear:

·         We suggest that product x,y,z will meet your requirements and we can validate it – but you can use anything that you want to (I want a predefined stack)

·         You have to pay extra for monitoring (I want a predefined stack)

·         You have to pay extra for stack capacity planning tools

·         To complete the integration you will need to pay $<Insert lots of money here> (why?? It’s a predefined stack)

·         Oh we don’t have full capability to monitor just yet (but it’s a predefined stack surely you can do that?)

·         We are working on capacity planning models (erm – it’s a predefined stack  - you know the characteristics and can plan workloads)

·         We are working on a roadmap – but we haven’t got one at the moment (erm – you know which components you are going to use, understand the roadmap (yep here it comes – it’s a predefined stack))

·         You don’t have to worry about that, as we take care of it for you – YEAH RIGHT!!!!

·         Every company can use cloud – you just need to plug our hardware in

·         Ahhh yes – we recommend this product – but it is not included as part of the overall cost of stack – you will have to engage with that vendor directly (erm – why…. It is part of the stack, surely, Isn’t it????)


In short – when you come and make this promise to me that a given integrated solution can reduce CapEx and OpEx, meet all of my business requirements, offer world peace and solve the current economic situation – please ensure you can back it up and the business case stands up.

I don’t expect to have to teach you about how the elements bolt together, know more about your infrastructure than you do, and really want the vendor to test my knowledge – not the other way around.

Finally – hypervisor technology is commodity – please remember and plan for this… Customer tie-in is not a good place, as soon as you do this; you are turning a commodity item into non-commodity. As an end customer – I want choice…

If I have to sit quality engineers into meetings to understand how to build kit into our infrastructure and spend a heap of time doing it – the vertical stack exercise has failed….