As I read, there are two patterns to define one microservice, by business capability and by subdomain. But I still find it very ambiguous. I get confused how these two patterns differentiate from each other. Both of them revolve around activities involving an area of business logic. All of components in each service are small enough to get packaged with each other without affecting other services. Could anyone please give me a further explanation about these two?
The commenters are right - there are some subjective definitions at play here. But there are some principles and concepts that can help reason about the different approaches.
Conway's Law
It's not strictly the original definition, but I think the distinction can be better understood with reference to Conway's Law:
Any organization that designs a system (defined broadly) will produce a design whose structure is a copy of the organization's communication structure.
The Inverse Conway Manouvre
Following from that thinking, the Inverse Conway Maneuver evolved:
The 'Inverse Conway Maneuver' recommends evolving your team and organizational structure to promote your desired architecture. Ideally your technology architecture will display isomorphism with your business architecture.
The Inverse Conway Maneuver is an attempt to structure your organisation to harness Conway's Law to achieve a better system design.
Decomposition By Business Capability
With an understanding of these concepts, we can consider decomposition by Business Capability to be guiding the system design according to the way the business is structured. This echos Conway's law.
The pro of this approach is it helps to ensure alignment between development teams and business structural units. The con is it may bake business inefficiencies that arose before an automated system was considered, into the design of your system.
Decomposition By Domain
Domain Driven Design (DDD) provides a suite of tools and methodologies to reason about the underlying domain at hand, to reflect the best available understanding of the domain in the software design and to evolve the software design as understanding of the domain grows and changes. DDD Strategic Patterns guide the creation of a Context Map which can form the foundation of your microservices decomposition.
From this, we can consider decomposition by Domain to be guiding the system design according to an analysis of the processes and information flows .
The pro of this approach is that it can lead to a system design that closely models the reality of what is happening (or needs to happen). Hopefully the business structure already aligns with this - but where it doesn't, it can reveal inefficiencies in the existing business organisational structure.
If you have influence over the organisational structure, this can be a foundation for utilising the Inverse Conway Maneuver and to allow you to evolve the software, the dev teams and the business units to achieve alignment.
If you don't, you may end up introducing friction points where the system design becomes misaligned with the business capabilities.
Conclusion
The reality is, neither approach is mutually exclusive - you will probably end up with a compromise that attempts to balance alignment with business capabilities as they are already understood and problem domains as they are revealed through a DDD process.
Business Capabilities do not reflect 1-to-1 the organisational structure. The represent what the business does (what capabilities it has) without specifying how they will be implemented. The organisational structure is part of the "how" / implementation.
If your business capability map is structured according to your organisational structure, then this should be evaluated more closely, unless of course, you create the business capability map first and then changed the organisational structure accordingly.
I've dealt with both Domains and Business Capabilities. They are essentially the same, depending on the level at which you are working.
Related
Im new to learn DDD concept and i cant understand something.
1-What difference between Context Map and Bounded Context and SubDomain?
2-How to recognize relation between Bounded Context ?
As said in the comment, this is a wide subject, and very important in DDD. It is the strategic part of DDD. Anyway I will try to answer your questions with an overall explanation:
DDD is about understanding and distilling the domain of the problem we want to solve. It is a continuous process of learning about the domain, talking to the domain experts. All people (developers, business people, etc) speak the same language. This language is used everywhere (conversations, documentation, source code, ...). It is called the Ubiquitous Language (UL).
The problem domain may have different areas of functionality, which would be domains too. They are the subdomains. So a subdomain is a subset of the problem domain. It is like splitting the problem into smaller subproblems, and a subdomain would be the domain of a subproblem. There are 3 kinds of subdomains:
Core: The point of distillation is to discover the subdomain that has value for the business, i.e., the one that will make your product better than others of the same kind. Such subdomain is the "core subdomain". For example, in "project management", the "task assignement" would be core.
Supporting: It is specialized in some business aspect that helps the core functionality. For example, in "project management", a "calendar" (for marking tasks delivery dates).
Generic: Functionality that maybe needed by any kind of application. For example, authentication and authorization of users.
Subdomains belong to the problem space.
To solve the problem, you model the subdomains, and you create bounded contexts (BCs). In practice, a BC is an autonomous application that contains the software model of a subdomain. A BC has its own UL. It is the context on which a term of the UL has a meaning. UL and BCs are the most important things in DDD. UL drives the BCs identification.
Ideally, you should align 1:1 the subdomains of the problem space with the BCs of the solution space, i.e., you should have a BC for each subdomain.
A team can develop one or more BCs, but a BC should be developed by just one team.
BCs belong to the solution space.
Context Map: It is a drawing that shows the BCs, and the relationships among them. Every relationship is classified in one of the following patterns:
Partnership
Shared Kernel
Customer-Supplier
Conformist
Anticorruption Layer
Open Host Service
Published Language
Separate Ways
Big Ball of Mud
Recognizing which pattern to apply in a relationship it will depend on the particular case you have. Some things that you have to consider are:
The 2 teams collaborate together.
One of the teams doesn't care about the other one.
The teams can negotiate.
The teams are independent.
Changes on a model (upstream) affects to the other model (downstream).
As #Augusto mentioned, this is a couple of chapters in the blue book, but here goes.
The domain model is found in the business rules and how people talk but a simplification of it is captured in code. Certain naming is consistent and the necessary invariants are enforced in the model.
A bounded context is mostly conceptual (might be a namespace, module, project in code as well...). It is the intention to keep a domain model consistent within it. So within the context, a certain ubiquitous language is used. And a model need only serve the needs of THAT context. It is the boundary in which the model can be used. In terms of recognizing these relationships? Some might be subtle but most are not. At least some people in the team will want to "avoid duplication" by unifying the model... so that is a clear indication that there is a relationship. Names are often the same or similar... or could be the same but one is better suited to one domain and another to another domain.
A context map is a bit more of a project management tool. It is a map of how different contexts (and the models within) relate to each other. In an Ordering Domain in an e-commerce system you may have a product. It would lead to A LOT of complication trying to have a unified Product in a model that spanned Ordering, Payments, Content for the website and Inventory domains (for example). So each of those domains should have a separate model. The context map is a diagram and related documentation that relates these bounded contexts together since there would be relationships and translation of data across from one model to the next, as an order flows through the system.
The last element you asked about is a subdomain. Here you probably are referring to a generic subdomain. Personally, I think the name is a little confusing. It makes it seem like a subset of the model. Maybe this is on purpose but I generally think of them as their own domain, just one that is not central to the business's proposition. For instance, if the aforementioned e-commerce company was known for its same day or next day delivery, then they probably shouldn't buy an off-the-shelf solution to inventory and shipping management. On the other hand, if they were focusing on a market that just wanted the cheapest deal but didn't mind waiting a few days, then that would be a perfect candidate for a generic subdomain.
My DDD glossary which has plenty of links at the bottom to more detailed articles.
If you are serious about learning this subject and can get your hands on some books:
Domain-driven Design by Eric Evans
Implementing Domain-driven Design by Vaughn Vernon
Domain-driven Design made functional by Scott Wlaschin (my favourite)
I'm wondering if System Sequence Diagram (SSD) belongs to design part or analysis part?
A System Sequence Diagram (SSD) is be a special type of UML sequence diagram that intends to document for one specific use case the the sequence of exchanges between the system under consideration and the outside actors.
It is not a standard UML diagram, but build upon such diagrams. The book "System analysis and design in a changing world" seem to have popularized this approach, but I could find articles dating back to the early 2000' (like this or this).
The above mentioned book places the SSD in the analysis activities. The reason is that analysis is about understanding the requirements, which often start with use-case. The SSD is then a fine-tuning of this analysis.
However, one could argue that it's part of the design activities, since the use case are the requirements, but how these requirements are addressed through a sequence of exchanges is already the start of the design of a solution, exactly as when you start to sketch an UI: more than one SSD could satisfy the needs and you have the choice.
So the answer depends on the purpose for which you're using the model.
My own point of view is that you're already drafting a solution, so it should be design, unless you do some reverse engineering of an existing application, or your client has very detailed requirements
Elaborating a little on Christophe's answer:
I would add that analysis and design are two highly intertwined activities, so you would probably see these SSDs in both contexts and it would be perfectly fine and acceptable. Use Cases, those that involve a system, are necessarily a design artefact (they are a design of what the system does in relation to external actors) although you can of course see that same thing as a pure analysis output (telling you what the system is required to do). These things are very hard to separate. The point may seem philosophical (it is somewhat), but it is useful to think in these terms.
When you see people creating "Login" Use Cases you can bet they already stepped into pure design, in other words: functional decomposition. In analytical terms the state of a User being logged in is a constraint on a Use Case, not a Use Case itself. Having a Use Case called Login therefore represents only a design choice (incidentally, if you see this in contexts where there is a division of responsibilities between the people performing analysis and design, then you'd do well to consider it an analysis fail: the analyst is essentially designing the system and that's not their role). Analysts sometimes use Use Cases to model layers of requirements that relate only to business processes, usually referred to as "Business Use Cases", that don't involve any system per se. But the origins of Use Cases from 20-odd years ago was in the system space.
What are DDD recommendations for inter-domain referencing design?
Should I try to connect them as "Matryoshka" (put one into another) or it is better to create upper-level "inter-domain" business service?
P.S. Crossing this smooth water, I was unable to find anything useful to read in the Internet, and have started thinking that for this kind of things exist better term than "inter-domain referencing"... Am I right?
DETAILS:
I have two models/business services.
Semantically first domain (A) is CRM with sell/maintenance process for our goods, second domain (B) is "design" data of our goods. We have two view points on our goods: from seller perspective and from engineer perspective.
Actually each model is effective ORM (Object-Relational Mapping) tool to the same database.
There are some inter-domain activities e.g. validations (e.g. sometimes we can sell things to smb. only if some engineering rules are valid).
From developer's point of view I have two clear possibilities (reference B in A or create new cross reference domain/service C ). But from designer perspective I am lost in understanding what kind of Business Service I have when I compose business logic from two different domains.
As far as I know, DDD has no strict rules for 'inter-domain' referencing. At the end of the day your domain model will have to reference basic Java or .NET classes. Or it may reference specialized date/time or graph library (aka 'Generic Domain').
On the other hand DDD has a concept of Bounded Context. And it has quite a few patterns that can be applied when you work at the boundaries of the system. For example 'Anticorruption Layer' can be used to isolate you from legacy system. Other integration styles can be used depending on how much control you have over external code, team capabilities etc.
So there is probably no need to introduce artificial glue layer if you just dealing with two subdomains in one Bounded Context. Might also be worth reading Part 4 of DDD book (Strategic Design).
UPDATE:
Based on the information you provided, it looks like you only have one Bounded Context. You don't seem to have 'linguistic clashes' where the same word have two different meanings. Bounded Context integration patterns are most likely not applicable to your situation. Your Sales domain can reference Products domain directly. If you think of Products domain being more low-level and Sales being high level you can use Dependency Inversion Principle. Define an interface like ProductCompatiblityValidator in Sales and implement it in Products domain. And then inject the actual implementation at the application layer. This way you will not have a direct reference from Sales to Products.
In addition to what Dmitry has already said...
I think of any code that crosses bounded contexts as application layer code. I would have that application layer code reference domain types from both contexts (and their repositories) but not have two domains reference each other. I think it's OK to have business logic in an application layer if it specifically crosses domain boundaries and is unit-testable.
If you really have a hierarchy, then it would be OK to have the the more concrete subdomain reference the more abstract domain. However, I would be careful if this causes you to need to have domain objects reference repositories of any type. Pulling objects out of of a repository is rarely a true domain concept. Referencing repositories is best done in an application layer that sits a layer above the domain model.
Of course this is all as much art as science. I'd try modeling a thin slice of your application a couple different ways and see what friction you run into with each approach.
I'm really new to DDD and trying to grasp some of the concepts.
Could someone explain me the idea behind Domain Modeling in DDD.
I have already gone through wikipedia explanation: http://en.wikipedia.org/wiki/Domain_model but still seems like there are some gray areas in my understanding.
Based on what I understood, domain modeling involves building a model around the business entities to express their relationships, express the entities that participate in the model etc..
Isn't this something that has been in practice always? in Object Oriented world, you model business entities into classes, objects etc.. and build the software around this.
What I do not understand is the emphasis Domain Modeling gets in DDD. Is it the same object/class modeling that you find in OO world, or is this something new to DDD ?
How does it differ from Object Oriented design/modeling?
Your answers are highly appreciated.
One distinction is that a "proper" implementation of the Domain Model Pattern in DDD is isolated from cross-cutting concerns.
For example, it contains nothing to do with databases or other persistence. Where it contains validation logic, it is business validation, not "does the name exceed the column length?" validation.
The idea is that the domain model encapsulates "the business" -- in business terms ("ubiquitous language"), to the extent possible -- and exposes relevant aspects of the business to "the program" without acquiescing to the needs of the software.
On the flip side, "the software" is concerned with IO, UI, and the like, but delegates all business logic to the domain model.
In principle, you can wrap your domain model up in an assembly and use it across multiple applications. When business rules change, as they do, you have one very logical place in which to affect the changes (because the model is a 1:1 or nearly-so representation of the relevant aspects of the business and is described in the same terms as the business).
The Domain in DDD does not need to be implemented in OO. In my experience an OO domain model is usually best, but there are very valid examples of situations in which it may not be.
You might implement a domain in rules, with a rule engine (example in The Netherlands where this is done for a large mortgage application). Or you might do it in a functional language. The essence is that your domain, in however fashion it is implemented, is isolated from what I usually call the technical aspects of your application (or, as the previous answer calls it, cross-cutting concerns, although I think there may very well be cross-cutting concerns within a domain). An isolating layer, which may be implemented using adapters, makes the domain as much as possible, even completely, independent from the technicalities. This layer usually leverages patterns such as Facade and Observer.
I am fairly new to this discussion but I HAVE to ask this question even at the risk of sounding 'ignorant'. Why is it that we now stress so much on 'DDD'. The more I look into 'DDD' the more complex it seems to make my application. Whereas modeling my domain with the database helps keep my application consistent across layers. Then I can use DAL Helpers such as SubSonic or L2S to easily access that model. What is so bad about this? Even in enterprise applications?
Why do we strive to create a new way of modeling our domain when we have a tried and tested one?
I am willing to hear from the purists here.
You can't sell an old methodology, because too many projects failed and too many people know the old methodology anyway. There has to be a new one to market.
If you're doing fine with the old way then use what works. Do pay attention to new stuff, as some really nice ideas come along. But that doesn't mean everything old is bad and stupid. Usually you can incorporate new ideas into the old models to a large degree.
There does come a time to make a move. Like I wouldn't do OOP with structures and function pointers. ;-)
This is actually a really excellent question, and the short answer is "you can." We used to do it that way, and there was a whole area of enterprise (data) modeling. In fact, the common OOD notations evolved from ERD.
What we discovered, however, was that data-driven designs like that had some difficulties, the biggest of them being that the natural structure for a data base doesn't necessarily match well to the natural structure for code.
OOD, to a great extent, derives from the desire to make it easier to find a code structure that has a couple of desirable properties:
it should be easy to think out the design
it should be robust under changes.
The ease to think out design comes originally from Simula, which used what we now think of as "objects" for simulation specifically; it was convenient in simulation to have software entities that correspond to the things you're simulating. It was only later that Alan kay et al at Xerox saw that as a more general structuring method.
The part about robustness under changes had many parents, but one of the most important ones among them was Dave Parnas, you wrote several papers that identified a basic rule for modularization, which I call Parnas' Law: every module should keep a secret, and that secret is a requirements that is likely to change.
It turns out that by combining Parnas' Law with the Simula idea of a "object" as corresponding to something that can be identified with the real world, you tend to get system designs that are more robust under requirements changes than the old way we did things. (Not always, and sometime you have to be crafty, as with the Command pattern. Most objects are nouns, thing that have persistent existence. In the Command pattern, the ideal objects are verbs, things you do.)
However, it also turns out that that structure isn't necessarily a good way to represent the underlying data in a relational database, so we end up with the "object relational impedance mismatch" problem: how to represent the transformation from objectland to database-land.
Short answer: if all you need is a CRUD system that allows users to edit data, just build an Access front-end to your back-end database (or use a scaffolding framework like you mentioned) and call it a day. You should be able to lop off 70% of your budget vs. a domain-driven system.
Long answer: with a data-driven design, what does the implementation of the business model look like? Usually after a couple years of building on new features to your application, you'll find that it's all over the place: tables, views, stored procedures, various application services, code-behind files, presenters/ViewModels, etc. with duplication everywhere. When you're having a conversation with the domain expert about a new feature they are requesting, you are constantly trying to translate from the business language into the language around your implementation, and it just does not translate.
What typically ends up happening is that you are forced to communicate with the business in terms of the implementation of the system, and the implementation becomes the "ubiquitous language" that the business and developers are forced to use when communicating. This has a wide range of consequences. The domain experts in the business start believing that they are experts in the implementation domain, and they start demanding features in terms of implementation rather than the business need they are trying to solve.
Also, you'll find that most data-driven implementations do not follow the "conceptual contours" of the domain, and the components of the system aren't very flexible in how they can be combined together to solve the problem, because they don't map one-to-one with concepts in the business model. When code isn't cohesive, changes and new features may require modifications all over your implementation.
Domain Driven-Design provides tools for making your implementation so closely resemble the business model that it's easy for everyone to speak the language of the business. It allows you to write "executable specifications" that test your implementation, but can actually be understood by your domain experts.