Loading ...
Sorry, an error occurred while loading the content.

Re: Agile and Enterprise Architecture ?

Expand Messages
  • marty.nelson
    I have recently come to the define Enterprise as meaning a high multiplicity of stakeholders: users, managers and VP s - often multiple sets across multiple
    Message 1 of 7 , Jan 30, 2012
    • 0 Attachment
      I have recently come to the define 'Enterprise' as meaning a high multiplicity of stakeholders: users, managers and VP's - often multiple sets across multiple Business Units, IT, Ops, Execs, compliance, audit, government regulators, etc.

      It's not surprising that there are 'uncovered constraints', especially when there's no clear line of sight to how the business serves it's customers and makes money.


      --- In extremeprogramming@yahoogroups.com, Jonathan Harley <jdharley@...> wrote:
      >
      > It sounds like a very naive development team to me. It also sounds like
      > the business partner wasn't very involved either.
      >
      > I'm not sure that an enterprise architect is required to fix this kind of
      > naivete but I suppose it's one blanket approach that could work. I would
      > think a little experience (like the one you outlined) would prevent the
      > members of that particular team from making that mistake again.
      >
      > Maybe some training in uncovering constraints would help too. No?
      >
      > Best - Jon
      >
      > On Sun, Jan 29, 2012 at 9:44 PM, Ken <kenmccormack@...> wrote:
      >
      > > **
      > >
      > >
      > > >> enterprise architecture (usually seen as BUFD)
      > > I see this mentioned a lot in relation to emergent design... and I wonder
      > > if the "architecture is BUFD" correlation is ever correct??
      > >
      > > EA is about strategic business alignment - making sure all architects are
      > > properly trained and understand their goals. BUFD only occurs when we
      > > design for a set of circumstances that are not envisaged by the business -
      > > this is also a misalignment issue that EA seeks to address.
      > >
      > > Agile proposes incremental design for features - however not all design
      > > decisions are easy to change, or as localised as feature design. We've got
      > > to know early if an all-pervasive component is going to be fit for purpose
      > > before we waste $$$ on a non-viable design path. A lot can often be known,
      > > broadly speaking, up-front. The business will know what volumes they'd like
      > > to achieve [otherwise the ROI may not stack up for them!]
      > >
      > > As a recent example, I joined a project where architects had got
      > > hopelessly lost early in detailed low level design and didn't think about
      > > non-functional issues at all - they had no oversight to keep their eye on
      > > the ball
      > > - Didn't know what load was likely [never asked]
      > > - Didn't guess load factor [10x isn't unusual variation]
      > > - Didn't do any load testing ["it's very hard to do"]
      > > - Didn't prove design OK [every iteration was getting worse]
      > > - Didn't realise until the end that the system hit 60% of expected
      > > performance
      > >
      > > The business sponsor knew or was able to estimate the answers to all the
      > > early questions - they were never asked. Here, EA would have avoided these
      > > basic architectural oversights and given them something more concrete to
      > > deliver. You could argue that any BDUF would have worked out cheaper than
      > > 'emergent inept architecture'.
      > >
      > > Kennos
      > >
      > > --- In extremeprogramming@yahoogroups.com, Steven Gordon <sgordonphd@>
      > > wrote:
      > > >
      > > > The global outsourcing approach is already abandoning ownership of any
      > > > innovation on the software development side. If there is any business
      > > > innovation, it will be before the outsourcers do their work.
      > > >
      > > > Therefore, this approach only works where software development is not a
      > > key
      > > > part of the business. If the software is just a commodity to the
      > > business,
      > > > why not just outsource it totally instead of just outsourcing the coding?
      > > > Then the outsourcer is responsible for the whole software development
      > > > process and the result, and therefore can improve and innovate (because
      > > it
      > > > is their key business).
      > > >
      > > > SteveG
      > > >
      > > > On Sat, Jan 28, 2012 at 10:01 AM, MarvinToll.com <MarvinToll@>wrote:
      > > >
      > > > > **
      > > > >
      > > > >
      > > > > Ram,
      > > > >
      > > > > Ford Motor Company was confronted with a similar challenge in 2005 when
      > > > > they elected to pursue global sourcing... and at the same time desired
      > > to
      > > > > have Valuable Software (or Working Software) agility.
      > > > >
      > > > > Obviously, these are orthogonal objectives. What emerged over a six
      > > year
      > > > > period was an approach characterized by the following assertion:
      > > > >
      > > > > "Proven Application Architecture, Design, and Implementation patterns
      > > are
      > > > > effectively communicated when linked to a reference implementation┬ů
      > > > > ┬ůso that you can focus on business innovation."
      > > > >
      > > > > The US Federal Government has a pilot implementation using the approach
      > > > > that emerged from Ford. It is called Pattern Enabled Development and is
      > > > > supported by an open source Framework Abstraction implementation called
      > > > > SOAj.
      > > > >
      > > > > Although the public documentation is sparse at this time... a technical
      > > > > writer was recently employed and we should see progress in that
      > > dimension
      > > > > over the next few months at http://PatternEnabled.com
      > > > >
      > > > > _Marvin
      > > > >
      > > > >
      > > > > --- In extremeprogramming@yahoogroups.com, Ram Srinivasan <vasan.ram@>
      > > > > wrote:
      > > > > >
      > > > > > Agile advocates delivering business value as frequently as possible
      > > (from
      > > > > > few weeks to not more than few months ). I have also read ( I think,
      > > Bas
      > > > > > Vode and Craig Larman, but not too sure) that starting with a small
      > > team
      > > > > > (or a small project ) is much more valuable than going with a Big Up
      > > > > Front
      > > > > > Design and taking on a large project. Mike Cohn talks about "a tracer
      > > > > > bullet approach" for user stories where the implementation of the
      > > story
      > > > > > cuts across all the layers associated with the story (just like
      > > cutting a
      > > > > > cake)
      > > > > >
      > > > > > My question is does enterprise architecture (usually seen as BUFD )
      > > add
      > > > > > business value (specifically frameworks like TOGAF, Zachman or DoDAF
      > > )?
      > > > > If
      > > > > > so how ? Also, can someone point me to any case studies / papers
      > > where
      > > > > > organizations have successfully implemented an EA framework in an
      > > agile
      > > > > > context ?
      > > > > >
      > > > > > Thanks,
      > > > > > Ram
      > > > > >
      > > > > >
      > > > > > [Non-text portions of this message have been removed]
      > > > > >
      > > > >
      > > > >
      > > > >
      > > >
      > > >
      > > > [Non-text portions of this message have been removed]
      > > >
      > >
      > >
      > >
      >
      >
      > [Non-text portions of this message have been removed]
      >
    Your message has been successfully submitted and would be delivered to recipients shortly.