tag:blogger.com,1999:blog-2120400492278270391.comments2023-08-23T06:57:04.599-07:00Enterprise Architecture - A Practitioner's ViewChris Birdhttp://www.blogger.com/profile/13436436994311245922noreply@blogger.comBlogger72125tag:blogger.com,1999:blog-2120400492278270391.post-34970747853995347542022-11-30T08:03:40.285-08:002022-11-30T08:03:40.285-08:00Very creaative postVery creaative postPressure Washing Camdenhttps://www.pressure-washing-service.com/us/power-washing-new-jersey/pressure-washing-camden.shtmlnoreply@blogger.comtag:blogger.com,1999:blog-2120400492278270391.post-41499080917026756442022-03-22T01:29:57.574-07:002022-03-22T01:29:57.574-07:00Hi great reading yyour postHi great reading yyour postTaya Pollardhttps://www.tayapollard.com/noreply@blogger.comtag:blogger.com,1999:blog-2120400492278270391.post-2801032009238979312014-08-28T13:12:50.079-07:002014-08-28T13:12:50.079-07:00I'll keep my perspective simple - at the start...I'll keep my perspective simple - at the start of this venture someone had an idea using forms, it wasn't thought through or engineered properly but everyone jumped into the process. The rest has been improvements based on trial and error in trying to manage their supply chain and reduce risks.RJNhttps://www.blogger.com/profile/04844361589615449655noreply@blogger.comtag:blogger.com,1999:blog-2120400492278270391.post-38828545105335840282014-06-02T06:43:11.979-07:002014-06-02T06:43:11.979-07:00Having railed against SOA for over sixteen years (...Having railed against SOA for over sixteen years (the first time was in 1998 with a programmer from KPMG) I think I now have a succinct statement of the problem. Firstly, as Chris states, every programmer thinks integration is just a matter of RPC. Which of course it isn't. The second problem is that we need to separate interaction from integration. Both need routing and transformation, which is why the marketing teams at vendors think they are the same problem. But interaction needs syntactic transformation and two way message routing whereas integration needs semantic transformation and one way message routing. That is why Microsoft no longer uses BizTalk for interaction but rather acquired Apiphany for the task. <br /><br />The principle is that you can never call out of a transaction in an enterprise application as it stops you scaling. If you call out you serialize the transaction and become dependent on another system. That is, you become flaky and latent. <br /><br />Or as the CTO of Amazon puts it, all enterprise services must be asynchronous and autonomous. <br /><br />So rather than use SOA, a marketing term, separate your enterprise architecture into exploitation of interaction (using a Resource Oriented Architecture if possible) and integration (using an Event Driven Architecture if possible). <br /><br />This will also help you avoid the horrific SOA mistake of making all integration into orchestration, thereby increasing the complexity of the enterprise a thousand-fold.<br /><br />John John Schlesingerhttps://www.blogger.com/profile/02564163777644343980noreply@blogger.comtag:blogger.com,1999:blog-2120400492278270391.post-1531832257941902432014-06-02T06:32:56.632-07:002014-06-02T06:32:56.632-07:00For me the bigger picture here is that an enterpri...For me the bigger picture here is that an enterprise application needs to separate completely the read/write (scratchpad) side of the data model from the read-only (historical) side. Chris is well aware of the pioneering work done by the global distribution systems over the last ten years as the ratio of 'looks' to 'books' went from 5 to 1 to 1000 to 1. The same is happening in banking. If a teller transacts, it is 5 to 1. If I do it over the Internet 50 to 1, if I do it on mobile 500 to 1 and if the Internet of Things (credit cards, fridges) does it 5000 to 1. So we are changing our product to have two data models, a transactional one and a reporting one. This is not data warehousing, the data is homogeneous. But the change is profound. The read only database is a source of time consistent bulk change. The read only database is not an interface at all, merely a side effect of doing the record keeping. <br /><br />JohnJohn Schlesingerhttps://www.blogger.com/profile/02564163777644343980noreply@blogger.comtag:blogger.com,1999:blog-2120400492278270391.post-79296811804358075312014-06-02T06:23:15.420-07:002014-06-02T06:23:15.420-07:00A related issue which is also of relevance to EA i...A related issue which is also of relevance to EA is the phrase "That is a million to one chance". We were working on the code to ingest 12 million financial transactions a day and I pointed out what I thought was an error. The programmer said 'but that is a million to one change' and I replied 'so we will get it 12 times a day'. The underlying truth is that everything you do in Enterprise Applications has to scale. This is as true of coding for race conditions as it is for coding for pre-conditions (invalid dates). <br /><br />JohnJohn Schlesingerhttps://www.blogger.com/profile/02564163777644343980noreply@blogger.comtag:blogger.com,1999:blog-2120400492278270391.post-51256174123998052392014-02-17T12:14:12.815-08:002014-02-17T12:14:12.815-08:00Johnny was an philosopher-poet-IT Architect. Also ...Johnny was an philosopher-poet-IT Architect. Also recommended is his song ''A Boy Named Sue", an allegorically story about naming your IT components.Terryhttps://www.blogger.com/profile/11811505003358361979noreply@blogger.comtag:blogger.com,1999:blog-2120400492278270391.post-17121032002783771942014-02-15T14:21:37.988-08:002014-02-15T14:21:37.988-08:00Interesting view, and that song is so appropriate....Interesting view, and that song is so appropriate. <br /><br />I've also used a vehicle metaphor for buying a company and then customizing there software to look like the legacy system. "If you want a Chevy, buy a Chevy - don't buy a Ford and change one piece at a time."<br /><br />Thanks for sharing.<br /><br />Anonymoushttps://www.blogger.com/profile/06075638406512335385noreply@blogger.comtag:blogger.com,1999:blog-2120400492278270391.post-56037413743598111102014-01-31T15:13:50.942-08:002014-01-31T15:13:50.942-08:00I got around to reading Roger's comment today....I got around to reading Roger's comment today. The issue isn't around the date routines, but more what happens when you hear the phrase, "That can never happen". That principle matters at the higher levels of EA as well as at lower levels of coding and development. Perhaps at the business/IT juncture, hearing business teams articulate we will never, "sell your information." However at some point they may. Absolutes are always tough.Chris Birdhttps://www.blogger.com/profile/13436436994311245922noreply@blogger.comtag:blogger.com,1999:blog-2120400492278270391.post-31578763309510226832013-12-19T20:49:49.562-08:002013-12-19T20:49:49.562-08:00Metadata Manager Enterprise Architecture
is a web-...Metadata Manager <a href="http://www.adaptive.com" rel="nofollow"><b>Enterprise Architecture</b></a><br />is a web-based repository that offers enhanced capabilities in the areas of data governance, regulatory compliance and metadata management.<br />Anonymoushttps://www.blogger.com/profile/14156927881909676131noreply@blogger.comtag:blogger.com,1999:blog-2120400492278270391.post-79801340384739672532013-04-09T22:27:24.167-07:002013-04-09T22:27:24.167-07:00Pretty good post. I just stumbled upon your blog a...<br />Pretty good post. I just stumbled upon your blog and wanted to say that I have really enjoyed reading your blog posts.<br />We combine extensive <a href="http://www.collaborative.com/consulting/enterprise-architecture-consulting/" rel="nofollow">enterprise architecture</a> consulting experience with your specific business needs and environment to help you determine the right architecture and technology.Anonymoushttps://www.blogger.com/profile/13314363083256903165noreply@blogger.comtag:blogger.com,1999:blog-2120400492278270391.post-79391559090070683332013-03-22T01:16:03.692-07:002013-03-22T01:16:03.692-07:00Should we also be interested in the rate of change...Should we also be interested in the rate of change of the rate of change? (Weak Signal).Richard Veryardhttps://www.blogger.com/profile/04499123397533975655noreply@blogger.comtag:blogger.com,1999:blog-2120400492278270391.post-37621187504819490432012-10-29T08:37:26.901-07:002012-10-29T08:37:26.901-07:00Certainly Richard, there are egregious examples of...Certainly Richard, there are egregious examples of people with an interest in the outcome making sure they influence the outcome. In the words (sort of) of Stalin and Beria. The outcome is not determined by the voters, but by those who count the votes.<br />In a children's book, I remember a character (a butcher) whose name was "thumb on the scales".<br />The bigger (from a systems persective) point is, I think, that we tend to view data at point in time, and whenever there are distributed systems, the time taken to synchronize state can always allow ambiguity to sneak in. If we are going to use eventually consistent models and the outcome depends on consistency, then we have to wait until "eventually" whenever that isChris Birdhttps://www.blogger.com/profile/13436436994311245922noreply@blogger.comtag:blogger.com,1999:blog-2120400492278270391.post-32227385136472899932012-10-29T07:04:41.834-07:002012-10-29T07:04:41.834-07:00Clearly history matters. There are many possible s...Clearly history matters. There are many possible sources of error and injustice in a voting system, but there are some specific tactics that have been used in the past to influence the result of an election, and these are the ones that the political parties tend to get most worked up about. Whereas most of the possible errors you identify would not obviously benefit one side more than the other.<br /><br />However, there is a second-order problem. If there is a significant number of uncertain cases requiring some human judgement, then this introduces the possibility of a biased official exploiting the uncertainty to swing the election.<br /><br />Fortunately, in a well-run democracy, you would never get the candidate's younger brother in a closely fought election controlling the administration of hanging chads. That would be like a referee in a <a href="http://www.bbc.co.uk/sport/0/football/20020062" rel="nofollow">soccer match</a> sending off two members of the Chelsea team and ignoring the behaviour of the Manchester United team. Such things just couldn't happen, could they?Richard Veryardhttps://www.blogger.com/profile/04499123397533975655noreply@blogger.comtag:blogger.com,1999:blog-2120400492278270391.post-16767205273343711872012-05-20T08:48:01.860-07:002012-05-20T08:48:01.860-07:00Richard I understand that you have only one (birth...Richard I understand that you have only one (birth at least) mother. But that relationship isn't exactly one to one. Without delving into family history completely, she could have many children (your siblings), and she existed before you did. So the role of Richard's Mother is, I agree in 1:1 correspondence, but the person who is Richard's mother is not in 1:1 correspondence with Richard or anyone else.<br />So, I guess it depends on which abstractions we use. I (at the risk of descending into a Clinton-esque land of definition) would argue that the word "is" is inappropriate (but convenient shorthand). It would be a bit cumbersome to say, "This person who plays the role of Richard's only birth mother" in conversation. So we use "is" in this rather corrupted form. I am sure the cunning linguists reading this can tell me what the linguistic term for such use of the verb to be is.Chris Birdhttps://www.blogger.com/profile/13436436994311245922noreply@blogger.comtag:blogger.com,1999:blog-2120400492278270391.post-26617570934884788142012-05-18T16:58:37.943-07:002012-05-18T16:58:37.943-07:00As it happens, I only have one mother, and she onl...As it happens, I only have one mother, and she only has one maiden name. Moreover, this is not going to change, ever. <br /><br />But when a service provider wants me to use my mother's maiden name as a secret password, I demur. So I have a collection of false mothers-maiden-names, Along with a collection of mythical first schools, imaginary first pets and so on. I then have the struggle of remembering which is which. <br /><br />By which time, I've forgotten why I needed to log on. Sigh.Richard Veryardhttps://www.blogger.com/profile/04499123397533975655noreply@blogger.comtag:blogger.com,1999:blog-2120400492278270391.post-45865672260765283612011-12-17T12:40:40.972-08:002011-12-17T12:40:40.972-08:00Chris,
A few thoughts in this context if I may. S...Chris,<br /><br />A few thoughts in this context if I may. Scalability is a property of the System. System is composed of application code and its infrastructure. While Clouds declare to provide "scalable" infrastructure the ultimate scalability includes application as well. So even if your cloud scales (whatever that means) your application in given cloud may scale or may not. Also, it may scale better if moved away from given cloud. The question becomes: are certain scalability features provided by given cloud aligned with my app current / future design and goals. Also, do I want to become dependent on certain scalability features of this cloud.<br />So to me as the System desinger the question is not only what scalability of the cloud is, but how it is realized and what additional technical constraints does it impose on my System.<br /><br />--<br /> Best regards,<br /> Konrad A.Konrad A.https://www.blogger.com/profile/14591329559313078782noreply@blogger.comtag:blogger.com,1999:blog-2120400492278270391.post-10876408459279867122011-08-17T21:09:57.101-07:002011-08-17T21:09:57.101-07:00Interesting - Zone no in your boarding pass anywa...Interesting - Zone no in your boarding pass anyway - so the number could be counted as well. Now how would you signal to your late arriving Exalted status person that they can "push in"...fickleshttps://www.blogger.com/profile/02297891327658894574noreply@blogger.comtag:blogger.com,1999:blog-2120400492278270391.post-89008055270488880592011-07-22T10:08:23.506-07:002011-07-22T10:08:23.506-07:00Of course it really needs to be the customer's...Of course it really needs to be the customer's voices that need to be heard. But that can be really hard to get. So maybe we have to have proxies for them. These proxies mustn't be the developers as such - they must be able to work in contrast to the developers.<br /><br />Architect interest - I think it is mostly in requirements that affect structure. Danger of course is that we get too esoteric. hang ourselves on the gallows of flexibility. But then if you are trying to build configurable products vs one off solutions, your axes of flexibility are different.<br /><br />I don't see much synthesis either - I wish I did!Chris Birdhttps://www.blogger.com/profile/13436436994311245922noreply@blogger.comtag:blogger.com,1999:blog-2120400492278270391.post-35400777212905915042011-04-28T10:26:40.436-07:002011-04-28T10:26:40.436-07:00Couldn't agree more. Only people rewarded for...Couldn't agree more. Only people rewarded for cross-silo work will work across silos.<br /><br />I wrote about this last year after presenting at the NIST think tank about "Collapse of the silo".<br /><br />http://viewyonder.com/2010/02/26/it-departments-and-the-collapse-of-the-silos/Steve Chambershttps://www.blogger.com/profile/10382328529822408184noreply@blogger.comtag:blogger.com,1999:blog-2120400492278270391.post-9165757905428349102011-04-14T07:07:46.695-07:002011-04-14T07:07:46.695-07:00@CharlesGary here. I was walking through the requ...@CharlesGary here. I was walking through the requirements forest picking mushrooms and came to this familiar grove. It is bright and shiny, but at odds with my reverie. I feel the need to synthesize a pint...<br /><br />A few things that give me pause:<br />1. BAs that synthesize to a solution-I don't see that much.<br />2. Architect's interest in requirements that affect structure-that seems reasonable, but what else?<br />3. Story telling-love that, but is it the BA's voice my team wishes to hear, or the customers?<br /><br />Just wondering...CharlesGaryhttps://www.blogger.com/profile/03974899781020455316noreply@blogger.comtag:blogger.com,1999:blog-2120400492278270391.post-79751726670139844182010-07-19T05:11:59.971-07:002010-07-19T05:11:59.971-07:00Aidan,
You said it well. There are definitely man...Aidan,<br /><br />You said it well. There are definitely many kinds of checking and reconciliation that should be routinely implemented in systems so that human trust is not undermined. I feel another posting coming along however - this time to do with systems of record and systems of reference, the CAP theorem, etc.Chris Birdhttps://www.blogger.com/profile/13436436994311245922noreply@blogger.comtag:blogger.com,1999:blog-2120400492278270391.post-23140920360448958182010-07-19T01:53:29.859-07:002010-07-19T01:53:29.859-07:00It seems that in the world of technical systems, t...It seems that in the world of technical systems, trust means mistrust. Don't believe what you are told about black box behaviour because it may well be wrong. On the news this morning it was reported that there was a vast increase in the level of fraud with the major frauds being committed by people who join their companies for the purpose. The level of due diligence for senior staff is lower than for mid-level staff.<br />However, mistrust has a high cost too. The previous comment could be used as an example: of course banks never trust their customers and the denial of their own venality is at the roots of the credit crunch and its unfolding sequelae. I think what you are saying, Chris, is that certain sorts of checking and reconciliation ought to be routinely implemented in systems so that systems don't cause cost by undermining human trust.Aidanhttps://www.blogger.com/profile/12795910135215132283noreply@blogger.comtag:blogger.com,1999:blog-2120400492278270391.post-86510699501943627752010-07-10T17:10:53.695-07:002010-07-10T17:10:53.695-07:00To me the story told by "Samples" shows ...To me the story told by "Samples" shows an interesting balance between the Policy/Value/Trust axes. It shows the need that even when you think you have a black box to "Trust but Verify" What seems "obvious" to Samples (and to most right thinking people, I expect) is clearly twisted in his/her example.<br />As architects/thinkers we are responsible for uncovering the trust boundaries and making them explicit. making decisions about how "optimistic" to be.<br />In this venal world, optimism is misplaced a lot of the time. <br />Just remember we have the best legal/government/political and other systems that money can buy!Chris Birdhttps://www.blogger.com/profile/13436436994311245922noreply@blogger.comtag:blogger.com,1999:blog-2120400492278270391.post-43010686320418070542010-07-09T12:37:50.461-07:002010-07-09T12:37:50.461-07:00Recently I underpaid my mortgage by $1. Recognizi...Recently I underpaid my mortgage by $1. Recognizing the error while balancing my checkbook and before the actual mortgage payment was due, I sent another $1 off to the mortgage company. Unfortunately their systems already went to "work" applying the underpayment as extra principal to the previous month's payment and applying the $1 payment as payment on the current month's payment, eventually making the current month's payment overdue. This took many weeks and phone calls to not only identify the problem but to get it corrected. Had I overpayed my mortgage by $1, my account would have been current and an extra $1 applied to principal with none of the ensuing people costs. Clearly the controls and trust here were only built in favor of the bank.Sampleshttps://www.blogger.com/profile/05614503312858735736noreply@blogger.com