Daniel Sholler
Research VP
13 years at Gartner
25 years IT industry
Daniel Sholler is a vice president in Gartner, where he advises clients on issues around application architecture, integration and development. Mr. Sholler is an authority on service-oriented architecture, and his current research focuses on… Read Full Bio
by Dan Sholler | January 15, 2009 | Comments Off
I have recently been inundated with questions about the design of message formats. While these questions vary from techniques for XML extensibility to issues about how to create common elements for messaging standards. Underlying all of these questions is a basic challenge: Creating a common set of shared services requires a common information model. We all have been challenged to create common information for many years, but because SOA is a centralizing concept (consolidating application specific functions into shared services) it creates more demand for sorting out the information model. The process of deciding how to structure common messages is very complicated, and tied up with all sorts of other design decisions about how services are structured, etc.
If anyone has good or bad experiences with this process of creating common message formats for SOA, please let me know, I would love to hear from you.
Also, if anyone is using industry standards as part of their message formats, I would love to talk to you as well.
Category: SOA standards Tags:
by Dan Sholler | December 23, 2008 | Comments Off
For you Gartner subscribers who are interested in WOA services, I suggest you take a look at the Reference Architecture for WOA. This is a description of the moving parts that are needed to implement a WOA-style service. The good thing about it is that there is not much there that is new; the architecture is completely consistent with that of the Web, and most of the content is derived from Roy Fieldings original thesis. Hopefully, it will be helpful to put that architecture into the service context. We are seeing a growing number of WOA-style services that are being deployed in organizations. In fact, our survey numbers indicate that while these services represent a very small percentage of those in production, 87% of the respondents had at least one WOA service in their portfolio. From the conversations I have had, it seems that many of these services were built in a Web Oriented way because the providers and consumers were intended to be web sites. I look forward to doing more research into how these services are built and used in the coming year.
A very happy holiday to all, and I look forward to speaking with you in the new year.
Category: SOA Tags:
by Dan Sholler | December 17, 2008 | Comments Off
OK, so I have finally dug out from the backlog after attending our AADI conference in Las Vegas last week. It was nice to see some of you in person. One thing that always bothers me is that I never have enough time to answer all the questions that people ask during our presentations. In this case, we also had a fire alarm go off during the Q&A portion of one of my presentations (it was a false alarm). So I figured that I would try to answer them here.
The link to my presentations is here.
One of the major points of discussion was that as you plan how to adopt and make use of the ideas inherent in SOA, you have to remain cognizant of the benefits that you are trying to achieve. One of the main benefits is "agility" which roughly translates into the ability to change our IT systems at a rate that is sufficient to match the needs and expectations of the business environment. One question that we received was about how to quantify agility. There is some Gartner research on this topic (Gartner subscription required) that goes into great detail on how to do it. One other heuristic that I have seen used in a few cases has been just to measure the cost of change, and the time to make changes. People who are doing this usually divide the changes up into small, medium and large buckets based on some arbitrary criteria, and then figure out how much money and how many person-days have been applied to making those changes. While this will probably not mean much for small numbers, in a large organization this can give very good directional indications of whether SOA (or any other initiative) is having an impact on your agility.
The other question that was asked was whether WSDL was required for SOA. This is easy, in that the answer is a resounding NO. However, WSDL is most often used to implement the particular style of SOA that we refer to as RPC style SOA. In this style, services are essentially methods, and the messaging network is assumed (from the point of view of the endpoints) to be a WS-* and SOAP based network. In this situation, there is a lot of tooling that people use to simplify the programming of the interactions, and in most cases that tooling relies on WSDL files. (even in this case, they are not required, but given the tooling and how easy it makes things, it is pretty unlikely that anyone would choose to do without. ) In other styles of SOA, such as the WOA style,(Gartner subscription required) WSDL is not only not required, but usually does not play a role. (Note that WSDL 2.0 was designed to be used even with WOA style services, and there are other efforts, such as WADL, to create a description language for this style of SOA. However, most practitioners at this point do not believe such a construct is necessary. )
Most of the other questions were from the presentation on the current state of SOA, which highlighted some of our findings from our SOA surveys. (Gartner subscription required). The first one was about whether SOA adoption was consistent across industries. The answer is that it is not, but that in some cases the number of responses in a particular industry was small, so it was hard to be sure the proportions were correct. In general, we see the most SOA adoption in high tech sector with 100% of the survey responses saying they were pursuing SOA, to a low of 37% in retail. However please be aware that the margin of error in this is too big to use these percentages directly.
The final question was about why, while most of the people using SOA saw positive outcomes for their agility and for other factors, there were some people who had negative or even strong negative outcomes. Most of the reason for this is just human nature.. if enough people do it, the outcomes will be all over the map. Anecdotally, a lot of the cases where people have had failures in their SOA initiatives have been the result of inattention to governance. As part of the conference, my colleague Paolo Malinverno did a presentation on the major mistakes to avoid, that is a compendium of the learnings we have had from talking to these organizations.
Hope you all enjoyed the conference. If you did not get a chance to attend, you can do so at the next one this summer in Orlando.
Category: SOA Tags:
by Dan Sholler | December 5, 2008 | Comments Off
I will be heading out to Gartner’s Application Architecture, Development and Integration conference next week at Caesar’s in Las Vegas. I am presenting on how to get started with SOA, how to use SOA effectively for integration, and the results of our surveys about SOA implementation and effectiveness. In addition, I will be hosting a panel on using standards with SOA, with an impressive group of industry luminaries who are involved in the standardization processes around SOA.
I hope to see you there!
Category: SOA Tags:
by Dan Sholler | December 3, 2008 | Comments Off
We all know that building the business case for many IT and infrastructure related projects is quite difficult. I and my colleagues Anthony Bradley, Roy Schulte, and Paolo Malinverno have summarized the best practices we have observed in a research note that was just published today . This technique represents a pragmatic approach to creating a business case for SOA, and probably could be applied to many other types of activities as well. Even if you are well on your way to establishing SOA as normal means of designing IT systems, it will be helpful to take a look at this approach and apply it to upcoming projects, as it enables you to identify those aspects of SOA that directly contribute to business performance. Once this type of technique is in place.. all sorts of things, from budget requisitions to measurement schemes, become much easier. It also helps in the never-ending task of justifying the value of the IT organization. I encourage you all to take a look.
Category: SOA Tags:
by Dan Sholler | November 7, 2008 | 1 Comment
One of the most interesting research projects that I have been working on this year has been our ongoing SOA user survey (Gartner subscribers only). Last week a summary of some of the interesting findings was sent out , and it has gotten some exposure (a search string). I am glad of that, because it has been quite a lot of work, but has thrown up a number of very interesting results. Much of the coverage has been focused in two areas:
1) There are leading signs that the adoption of SOA as a practice within organizations will slow next year. The proportion of organizations saying that they were planning to adopt SOA designs in the next 12 months declined, and the proportion saying that they had no plans to do so grew. This does not mean that organizations are abandoning SOA, but it does mean that an increasing number of those who have not done anything about it yet are not going to do so in the next year. This may partly be a result of saturation, but it is also probably related to the expectations of flat or shrinking budgets in a down economy, and the fact that many organizations have other fish to fry. Extensive analysis of these trends can be found (for Gartner subscribers) in the 2008 SOA User Survey: Adoption Trends and Characteristics note.
2) Another thing that is evident in that note is the worldwide distribution of the use of SOA. We found that European countries (Mainly UK, France, and Germany) were the furthest along in the adoption of SOA, while Asian countries (mainly India, China, and Australia) had much lower rates of adoption. Again, there are a number of factors here, but it seems clear that this imbalance is likely to remain for some time.
Just last week we published the 2008 SOA User Survey: Justification (Gartner subscribers only) research, which shows how organizations are justifying the organizational changes and investments needed to implement SOA. Among the results was a clear indication that the formal justifications that people use (which were often cost based) and the informal expectations of return (which were based on values such as agility measures) were not aligned. Part of the challenge with SOA may stem from the fact that expectations set at the time people justify their investments may be in conflict with the actual goals they are using SOA to accomplish.
This research is an ongoing project. I am currently working on an analysis of value capture from SOA based on the survey data. I will be presenting all of these results at our upcoming Gartner Application Architecture, Development & Integration Summit taking place December 8-10, in Las Vegas, NV. Please come, and if you do attend, it would be terrific if you could spend some time participating in the SOA survey. There will be survey kiosks at the event in the Gartner Realtime Research Center. Please come and visit, and I hope to see you there.
Category: SOA Tags:
by Dan Sholler | October 27, 2008 | 3 Comments
On a friend of mine’s social networking site, he has many pictures of his kids. Recently, he posted one of his son goofing off in (what I think is) his religious school class at church. It occurred to me that kids think that they have an advantage now, because they can chat online without their dorky parents overhearing them. (I remember taking the cordless phone into the bathtub to try to avoid being overheard) But social-network aware parents can use the power of the Internet to keep the kids in line as well. After all, I have lots of photos of my kids that they will no doubt find embarrassing when they are teenagers, and I know how to put them places where their friends can find them…
it seems logical to me that the discourses that we have in public in our daily lives, includes those that involve disciplining our children, will be replicated in the online social networking world. Or more to the point, if they are not, then it will be very clear that this world is only a limited representation of the social interactions that we have. So, I am anxiously awaiting more goofing-off kid pictures. On the other hand, most of us try to keep that activity confined to the places where it is really necessary. After all, it is not always the best idea to have that kind of conversation in public. As we use these kinds of sites and the line between public, between friends and private gets more difficult to police, there will no doubt be many familial challenges about what was posted on Facebook.
I can see it now in the latest parenting handbooks:
Chapter 5. Rewards and punishments: When to bribe with candy and when to “out” them on myspace -
Category: Social networking Tags: child rearing, children, Social networking
by Dan Sholler | October 22, 2008 | 2 Comments
Our SOA surveys show that the vast majority of services in production today (over 80%) are WS-* services. This means that they have a WSDL descriptor, and at least in theory they can be invoked via SOAP over HTTP. In reality, many of them use commercial middleware or ESB message transports. However, nearly every discussion of SOA today includes a conversation about how much one can rely on the WS standards, and their evolution. Organizations of all types are trying to figure out if the vision of web services that was espoused a few years ago has any relevance to their future plans.
There are several reasons why this comes about. One of the most prevalent ones is that there is a growing understanding that SOA is not a uniform thing. There are different styles of SOA, and the differences between those styles are part of what creates this uncertainty. The prevailing style, which unfortunately does not have a recognized name, (although we refer to it in our research as "RPC" style SOA) is closely associated with the use of WS-*. An alternative style (Web Oriented Architecture coined by my colleague, Nick Gall) which uses the architectures associated with the web, and with the architectural style called REST as the basis for the design of services. This style is often perceived to be Anti-WS, since the WS-* protocols do not lend themselves easily to the creation of WOA or RESTian interfaces. The notion that there is an alternative, and one that does not depend on web services standards, and in fact may be antithetical to the use of those standards means that the position of those standards as the core of SOA understanding has been disrupted. WOA style services represent a small (about 15%) but rapidly growing segment of the services that are being built today.
In addition we are at a point where people are really doing SOA. In North America and Europe, the use of SOA is extremely strong (See our recent survey research). We also see from those same sources that most services are being implemented for A2A within the organization (although the use of B2B services is growing. The data for this is available in my upcoming research on justification… stay tuned). Many of these services are being built with the help of an ESB or other middleware technology forms, and while WS* may be used in those implementations, it is often hidden in the tooling that organizations use.
There is no doubt that WS-* based services will have a place in the service portfolio for a long time to come, but we are reaching the point where WS* does not mean SOA. This is both good and bad. The good is that it means that practitioners must now focus on the design, which is the part of SOA that really makes the difference. The bad news is that it means that the technology choices are not nearly as simple as they might have first appeared, and we will be struggling with choosing the right technologies for the design patterns we choose for a long time to come. In many cases, we will find the WS-* standards and the technology that implements them to be useful, but we are likely to find any number of cases where it is not. As we continue to evolve the styles of SOA, we will have more variation in the manner in which services are implemented, at least until the industry as a whole comes to a common understanding of the best practices and uses for the various styles of SOA.
Category: SOA standards Tags:
by Dan Sholler | October 17, 2008 | Comments Off
Having spent 4 days talking to dozens of our clients about SOA, it is impressive how far we seem to have come in such a short time. While there are certainly any number of "how do we get started" questions, many of the people I am speaking with are well on their way to implementing SOA system components.
Nearly all of the questions revolved around the fundamental issue of how to determine what services should be created and what they should look like. This issue, which is often discussed under the heading of "SOA Governance", "SOA development methods" or "the role of the COE", shows that (at least for the folks that came to Symposium) we have moved beyond the idea stage and are focused on practical application of SOA.
When you think about the answers, this all comes back to the fundamental notion of shared services. Shared services are agreements; the users of a shared service agree that they are doing the same thing, and that it would be beneficial to do it in the same way. The implementation of that service can then be consolidated into a single shared service entity (whether an organization, or a module of an IT system) and the participants in the agreement will use that implementation. This is how business-level shared services (such as consolidated purchasing) work, and it is also how shared application services (in the SOA sense of the term) should work.
The process of service governance is to make sure that these agreements can be created, formalized, implemented and modified. There are many techniques and technologies that may help with these efforts, especially as the numbers of services (and therefore the number and diversity of parties to the agreements) grows, but first and foremost it is important to insure that agreement can be reached.
The problem of which services to start creating is essentially the same. Organizations must begin with the services over which there is already some agreement about their similarity, or at least the need for similarity. This may account for why many of the SOA projects that we see are essentially hand-crafted MDM. Master data is master data precisely because it is shared, and there is a recognition that the more similar the master data is across systems, the better the systems will be.
All of this backs up the information from our surveys, which has shown that service governance is the most difficult aspect of leveraging SOA. If your organization is trying to move to SOA, you need to be sure to focus on the the fundamentals of service governance, and agree to agree.
Category: Getting started Governance SOA Tags:
by Dan Sholler | September 26, 2008 | 1 Comment
In the course of an interesting conversation on a different topic, I heard a rumor that could have a significant effect on the perception of open source. The person I was speaking with mentioned that certain agencies in the US Department of Defense and Intelligence communities were thinking of biasing their procurement policies toward open source, because they felt that it was more secure than traditional closed source products.
Now, the idea that the Federal government favors open source is not strange, mainly because it gives them flexibility that is constrained by their procurement rules (an example) . And furthermore the idea that they might be modifying the rules to favor open source is not strange (although the source of this rumor was someone who makes money from open source software, so they have a vested interest in the outcome). However, the idea that the organizations whom we delegate the responsibility for our national security have a reason to believe that open source is more secure would be huge.
The team here at Gartner is already seeing a tremendous uptick in interest in the past few months. The majority of our middleware product related inquiries include a discussion of open source alternatives. However, it is still usually the case that the strategic, mission critical projects use proprietary software, and the open source is used in a similar function for the rest of the portfolio. One of the main reasons for this is a lingering unease about the quality of open source products and the community development processes that create these products. If this rumor were true, that would go a long way toward dispelling that unease.
Whether or not this rumor is true, I expect to see a number of this type of event over the coming year that will substantially improve the way open source middleware (and open source products in general) are perceived.
(Gartner subscribers: check out the discussions of open source in government and open source in MOM
Category: Open Source Tags:
