<<It is no secret that a large percentage of Oracle Fusion Middleware
customers are using Oracle SOA technologies for integrating enterprise
applications such as Oracle CRM or ERP. Many may view these projects as,
(or maybe replacing), EAI like point-to-point integrations of legacy
systems to satisfy immediate business needs while others have charted
out a long-term SOA vision for service reuse. This illustrates some key
differences in various customers SOA adoption patterns which became
apparent to me during a recent discussion with a customer who is long
term user of Oracle BPEL PM for integrating various enterprise back end
systems. They were very familiar with the newer SOA tooling but when I
quizzed them on how they managed their service portfolio to achieve
shared reuse, I got blank stares which indicated to me they were not as
far along the SOA maturity curve as what I would have expected for a
long term Oracle BPEL PM customer. This lead to a detailed discussion
about the many benefits and challenges of adopting SOA and shared
services reuse strategy.
The benefits of this approach while evident to many were not clear to
this customer and I suspect many others as well, especially those in the
depths of complex enterprise system integration projects. Significant
savings can and should be realized when adding additional consumers to
that CRM endpoint. What SOA tooling delivers here is that classic ESB
many to one mediation consumer pattern that allows multiple consumers to
access the same endpoint by (re) using a SOAP or REST based proxy
service. Beyond the benefits of de coupling the service endpoint
provider from the consumer for ease of endpoint change management, this
proxy service can also be reused by portals, enterprise or legacy
systems for future integrations with other business units. In a
simplistic view of savings, it is easy to envision a 50% savings on
future integrations of that endpoint minus whatever overhead is
associated with governing and extending the service to meet additional
consumers needs. This infrastructure overhead should not be
underestimated and may take time but is key for reaching the critical
goal of building trust with new prospective consumers of these services.
So what does this all have to do with governance you may ask.
One of the main values of SOA Governance is to *deliver savings by
enabling this reuse of existing services* and underlying assets, by
creating and managing a shared service and asset portfolio. The first
goal is to deliver visibility of these services and assets since
logically, if you don't know what is available, then you cannot even
begin to reuse it. Once there is adequate visibility, there needs to be
some controls put in place to monitor and manage the usage of these
services to guarantee meeting the requirements of the various business
organizations using them. A hidden benefit of this approach is to
*promote communications among these various business stakeholders* but
that is for another topic. Good governance tooling can promote this
*Visibility, Control and Monitoring* necessary to manage a shared
service portfolio. This is loosely referred to closed loop governance.
The 2 main components for implementing closed loop governance are:
1) *Enterprise Repository* which provides the visibility and is the
single source of truth for consumption of assets such as WSDL, XML
schema's, BPEL processes, web service endpoints etc...
2) *Enterprise Monitoring* Manager enables IT staff and prospective
consumers the ability to see how a service is performing and key to
building that consumer trust we mentioned previously.
The Repository and Monitoring tools must be integrated such that when
consumers browse the repository, they can see historical information to
help them decide if this service will meet their requirements for
performance, security and availability.
While all of this may seem like SOA Governance 101 for service reuse, it
really is just meant to bring attention to the fact that many users of
SOA Technologies especially those doing point to point like integrations
with SOA tooling may not be thinking in terms of reusable service
portfolios. Even though they have built some services that are
potentially sharable, by adding some initial SOA reuse methodologies and
governance tooling, organizations can head down the path to realize
paybacks on their existing enterprise integration investments. As it
turns out, it is not a stretch to draw the modern corollary that *reuse*
is a very ecological, even *green* activity that is far more useful than
it's older sibling, recycling which requires additional investment for
repurposing existing materials. More on this later.>>
You can read this at: http://blogs.oracle.com/governance/
Gervas
