We've Moved...
The sun has set on Chaotic IT. I am now posting on The SOA Coach blog. Same content, stickier name. Hope you can drop by.
Defy. Change.
The sun has set on Chaotic IT. I am now posting on The SOA Coach blog. Same content, stickier name. Hope you can drop by.
65 million years ago something unexpected happened that altered the course of Earth's history forever. Dinosaurs had reached the apex of evolutionary fitness and had ruled over land-based animals for millions of years. They had become indomitable masters of their surroundings. Then, suddenly, in a geological instant, they vanished.
"They are rigid hierarchies through and through; it’s in their genes; it seems they are not and cannot be agile."If this is indeed the case, then these companies' existence is endangered. Today's business environment increasingly favors organizations that can self-organize quickly and dodge the effects of constant change. Heavyweights cannot maneuver around change, so they become big targets that sustain constant damage. Lacking the ability to adapt, time will eventually erode the pedestals out from underneath them.
I have been preoccupied for the last several months with traveling and becoming a new dad, but I managed to cobble together an article for the latest issue of SOAWorld magazine. It presents and contrasts two rather opposed SOA strategies--Mainstream SOA and Long Tail SOA--and argues that organizations embracing SOA tend to focus too much energy on building Services for reuse. Highly reusable Services perform business functions that are in high demand, hence the "mainstream" label, but these Services are scarce compared to the accumulated demand for one-off Services. In general, only 20% of a company's Services are highly reusable. Targeting only these neglects the other 80% of the company's Service market. To truly become agile, IT organizations should build Service portfolios at the Long Tail end of the spectrum, because Services do not need to be reusable to be valuable. . .they just need to be available.
I spent the first half of last week in New York city at the SOAWorld 2007 conference, where I had the privilege of presenting my thoughts on the ROI of SOA. It was a fun session with a very inquisitive audience. I have received a lot of valuable feedback on the presentation and the supplemental article in SOAWorld magazine and would like to address one topic that has come up frequently, and that is whether simply choosing SOA over point-to-point (P2P) integration is enough to achieve agility.
Just when I thought I had sold the SOA community on my perspectives on the ROI of SOA, Alastair Bathgate, a fellow tech blogger whom I respect, threw down the gauntlet and took some jabs at my recent whitepaper. Two thoughts filled my head as I read his retort:
"Although, as Marc points out, the cost of connecting all the components in a P2P architecture is undoubtedly high, I wonder if that would ever happen in reality..."In my paper, I based my ROI analysis on the assumption that total infrastructure connectedness is required to achieve maximum agility. However, it would be foolish to think that every last component in an IT environment would need to be connected to every other in order to produce appreciable return on integration investments. In fact, it may be altogether impossible. However, I do contend that business-critical segments of the infrastructure can be configured into fully connected integration sub-networks.
"[Marc's] target of zero latency...is again appealing, but unlikely to be achieved...even with the most sophisticated SOA."Again, he's right on the money. No company can ever hope to operate in absolute real-time, even with a perfectly managed service-oriented infrastructure. There are countless forces at play constantly that contribute friction to even the most streamlined business processes. (It is these forces that produce the chaos that is the theme of this blog.) In order to home in on the essence of SOA and P2P, though, I had to eliminate these variables from my equations, otherwise my analysis would have been so muddied up with extraneous data that the fundamental properties of P2P and SOA wouldn't have shined through.
"...the enterprise is always dreaming up new business processes...which means that maintaining a Service Oriented Architecture can be just as painful as building P2P solutions."SOA certainly can be just as painful, or more, than the P2P approach (I assume that by "painful" Mr. Bathgate means costly), but this is a function of the quality of a particular SOA implementation and is not an immutable property of SOA itself. Some SOA initiatives will succeed and some will fail. That's just the way it is, but we shouldn't fault or avoid SOA just because success is not guaranteed.
"...the orchestration layer of any SOA needs constant reconfiguring to meet new business requirements."
"I think that Marc has written a highly interesting and thought provoking paper that considers the “pure” arguments. He points to a fork in the road where a decision must be made to take the P2P or SOA paths. I just think that in the real world there is some dirty middle ground that we have to accept exists..."Again, I could not agree more. My ROI model is a starting point for understanding the raw characteristics of SOA as compared to P2P. It is not meant to be construed as a methodology or formula for success. It places SOA and P2P into a vacuum, free from the influences of external, real-world forces, systematically dissects them and analyzes their DNA under a microscope. Indeed, it is a highly purified view of two extreme approaches to enterprise integration. Perhaps overly scientific for many practical purposes, but necessary to reveal their most fundamental characteristics.
Labels: roi soa
Once again, Joe McKendrick out at ZDNet published a very quotable blog article. He cites a few notable sentiments from Gartner research director, L. Frank Kenney, about the outlook of SOA and throws in a few of his own.
Labels: soa roi gartner
One of the most difficult aspects of Service Oriented Architecture (SOA) is demonstrating tangible return on investment (ROI). Many folks concede that SOA is more of a long term value proposition than traditional architectural approaches, but that doesn't do much to lower the blood pressure of the business sponsors footing the bill. SOA programs need strong business backing to be successful, so SOA advocates need to be able to convince the business, and keep them convinced, that it is the right way to go.