I am trying to understand the definitions in this document.
http://www.opengroup.org/soa/source-book/ontologyv2/service.htm
Their definitions of service, service interface and service contract are either unclear or seem different from what I normally encounter.
Service:
“A service is a logical representation of a repeatable activity that
has a specified outcome. It is self-contained and is a ‘black box’ to
its consumers.”
Lets say I have a WCF project and it has two Operations
StoreFront
+GetPrice
+AddToCart
The definition says "a repeatable activity". So is the service StoreFront? Or do I have two services (GetPrice and AddToCart).
Service Contract:
Has an "effect" class. Is the effect "return price" and " added to cart" ?
From the same article:
“A capability offered by one entity or entities to others using
well-defined ‘terms and conditions’ and interfaces.” (Source: OMG
SoaML Specification - my italics)
This is in my opinion a preferable defnition than the one talking about "repeatable activities".
The key word in the definition is capability. Capability refers to Business Capability which is a carry-over from the BPM industry, but in an SOA context refers to a business domain with distinct boundaries.
So from this definition we can surmise that services should be exposed or should operate within a business capability/process boundary. This leads us towards the idea (from the principals or tenants of SOA) that services should be autonomous within well defined boundaries.
In your example, you are asking
So is the service StoreFront? Or do I have two services (GetPrice and
AddToCart)
The answer to that as always is "it depends". However, generally Pricing (GetPrice) would belong to a different business capability to Ordering (AddToCart). Additionally, the operations differ in some other important ways:
GetPrice is a read operation, while AddToCart is a write operation.
GetPrice is a synchronous operation, while AddToCart could very well be asynchronous
So from these we should probably assume that they are two different services from a business perspective.
This assumption has some radical repercussions. If they are two services, then according to SOA they should be autonomous. Meaning that we should be looking to minimize coupling between the services in every possible way, so that as much as possible they can be planned, developed, tested, built, deployed, hosted, supported, and managerd as separate concerns.
Another repercussion is that when you physically separate services to this extent, how can you show this stuff together to your users? They may be different capabilities but they still need to work together on the screen.
Additionally, from a back end perspective Ordering needs to know about Pricing data, otherwise how can order fulfillment happen? If you've separated the database into two, how can the Checkout service know how much stuff costs, what discounts to apply, etc?
I have posted about this stuff before, so please feel free to have a read. I would recommend reading the excellent article on Microservices by Lewis and Fowler also.
I'm reading all over the net that you your separate your "external schemas" from your "internal schemas" and never expose the "internal schemas" to any external actor.
If my solution only acts as a messagebus to create a loose coupling between 2 existing systems, will I really need any internal schemas?
System A makes a Request(Message with SchemaA) to Biztalk
Biztalk Maps SchemaA to SchemaB
Biztalk forwards request of type SchemaB to SystemB
SystemB returns ResponseB
Biztalk maps ResponeB to ResponeA
Biztalk routes the result back to System A
I can't see the pro's of having an internal schema and map:
SchemaA -> SchemaInternal -> SchemaB
?
The term canonical schema is often used to describe the creation of schemas internal (SchemaInternal in your last example) to an integration mechanism such as BizTalk.
Use of canonical schemas is widely regarded as a best practice, as it decouples your BizTalk flow control mapping from any 'other' system's schemas (other system here could be internal to your organisation or external to it, e.g. a supplier, customer or partner system). This way, if any of the systems integrated via BizTalk change, it is just the external schemas, and maps to the canonical schemas which need to be changed. It also prevents foreign conventions, naming and hierarchy differences inherent in external schemas from leaking into your internal BizTalk artefacts.
Generally, transformation of incoming messages to a canonical schema is done as early as possible e.g. on a receive, and similarly, transformation out of canonical done as late as possible, e.g. on a send port map.
A common scenario for Canonical Schemas (CS) is where a single orchestration or message flow is common to multiple trading parties (e.g. you may have many suppliers with different systems, however, all of them submit invoices for processing). In this case, each new supplier system just needs to be integrated with your CS - no new processing logic needs to be added or duplicated - CS can actually reduce the overall effort in such instances. (The n x m problem is explained in detail here). Another example of where CS are vital is where your business IS switching of messages - e.g. a Medical industry switch will have many doctor and practice systems sending authorisation requests and invoices and these need to be mapped and routed to multiple medical fund (medical aid) systems.
And FWIW:
IMO CS make most sense in an when BizTalk is the end-end solution in an EAI or ESB scenario, e.g. direct integration of 2 or more line of business systems. Otherwise, if BizTalk is just one endpoint on a larger corporate ESB, then it probably makes sense to use the corporate ESB schemas internally, and hence map external schemas directly to the ESB schemas (i.e. no need for another set of CS within BizTalk, provided that you have a good change management / version control mechanism across your enterprise).
If standard schemas (e.g. EDIFACT) exist for your industry, it is moot as to whether it is a goal to adopt these as internal CS. In general these may conflict with the meaning of Canonical as being 'simple', as industry schemas often need to be verbose in order to model all flavours and 'edge cases' of the document). Personally I would ensure that I have a mapping to / from said industry schemas, but would use a custom schema internally.
In described solution you don't have need in internal schemas. Well you can hide the schemas of System X from users of System Y, but that is not so important.
In this context, External = Public, meaning outside your organization.
The guidance is to protect internal implementation details, naming conventions and such, from others.
If both System A and System B are inside your organization then 'security' is less of an issue but your application can still offer an 'external' schema to consumers in order to protect them from internal changes to your application.
I've been reading up about ID-based encryption, but can't find any implementations (the maths in the paper is a bit much for me). Does anyone know where I might go looking to find an implementation? I don't really mind what language it's in.
You can try Stanford PBC or IBE toolkit
IBE permits any string of bits to be used as a public key, and the private key is derived on the fly from it by the PKG after authentication. Private key generation in this manner is not possible with methods like RSA or using classic CA's where pre generated key pairs are needed before any encryption can take place.
IBE is a different approach and avoids the need for the certificate to bind the public key to an identity: authentication and identity verification is independent and separater in an IBE scheme versus tightly coupled in legacy PKI. This means that the need to generate, bind, manage, publish, revoke, have responders for etc for public key certificates is removed. Also, given all keys are generated other than the root key(s), there's no database of public keys, no need for an escrow database etc. This makes operational matters far simpler.
Generally IBE keys are short lived too e.g. a week, an hour etc which removes the need for a revocation process as they naturally retire.
Given the public key can be any arbitrary string, this means that time can be bound into the string, as well as policy information. This provides a lot of flexibility to handle ad-hoc public key encryption and naturally permits federation - very important in cloud environments for example.
Commercial implementations exist - www.voltage.com being the most well known. Its standardized - IEEE 1363.3 and IETF 5091 et al.
http://www.ietf.org/rfc/rfc5091.txt
SSL with certificates is a form of IBE. Verisign is the PKG from the article; it generates certificates for verified and trusted network channels, which expose the RSA public key used to send messages to this endpoint. The private key is also obtained via the PKG. Now, when communicating with this endpoint, callers can encrypt with the public key from the word go, meaning no information has to be sent "cleartext".
Closed. This question needs to be more focused. It is not currently accepting answers.
Closed 9 years ago.
Locked. This question and its answers are locked because the question is off-topic but has historical significance. It is not currently accepting new answers or interactions.
Can someone explain in plain english what is SOA all about? I hear SOA here, SOA there but I cannot understand exacly what it is and what is used for. Was it some simple concept and later evolved into something huge or what?
All documents, including wiki are a bit abstract or maybe I'm an idiot and don't get it. Is there an idiot's guide on this?
What exactly is there behind these three letters?
SOA is a new badge for some very old ideas:
Divide your code into reusable modules.
Encapsulate in a module any design decision that is likely to change.
Design your modules in such a way that they can be combined in different useful ways (sometimes called a "family" or "product line").
These are all bedrock software-development principles, many of them first articulated by David Parnas.
What's new in SOA is
You're doing it on a network.
Modules are communicating by sending messages to each other over the network, rather than by more tradtional programming-language mechanisms like procedure calls. In particular, in a service-oriented architecture the parts generally don't share mutable state (global variables in a traditional program). Or if they do share state, that state is carefully locked up in a database which is itself an agent and which can easily manage multiple concurrent clients.
You might find this article (What is SOA? - SOA and Web Services Explained ) helpful.
A little teaser:
SOA is a style of architecting applications in such a way that they are composed of discrete software agents that have simple, well defined interfaces and are orchestrated through a loose coupling to perform a required function.
There are 2 roles in SOA- a service provider and a service consumer. A software agent may play both roles. SOA is not an entirely new concept – however, this article mainly focuses on SOA as implemented with web services.
I see many answers explaining a Service Oriented Architecture (SOA) using even more advanced words and technical terms. I'd like to give a shot at explaining it for the layman, using an analogy in plain english.
But first a description of a SOA
SOA could be described in three layers as seen in the picture below. On one side we have the Provider and on the other side we have the Consumer, separated by a Bridge where the two sides communicate.
The consumer uses a number of Applications necessary for it's business and the provider uses Components that provide these applications with information. They communicate through a set of Services using a common architecture.
The analogy
Imagine a house on the country side, that in many ways is part of a larger community, like a city or town. The city has it's own complex systems for providing water and electricity, handling sanitation, providing transportation and other utilities. The House is the consumer in this model, the City (or community) is the provider and the pipes, sewers, powerlines, optical fibers etc. is the Infrastructure in which they communicate.
This model could loosely be compared to a SOA. The people in the house uses a number of different "applications" like radiators, computers, toilets, lamps, underfloor heating, bathtubs etc. These applications don't care how the city generates the water, creates the electricity or handles the waste as long as it works. The components of the city are generators, water pumps and sanitation areas. It provides the house with all these needs but it's up to the house to use it in what ever way it sees fit.
I hope this gave at least someone a better picture of a SOA.
Let's assume you have four cooks. In SOA, you assume they hate each other, so you strive to let them have to talk to each other as little as possible.
How do you do that? Well, you will first define the roles and interface -- cook 1 will make salad, cook 2 will make soup, cook 3 will make the steak, etc.. Then you will place the dishes well organised on the table (so these are the interfaces) and say, "Everybody please place your creation into your assigned dishes. Don't care about anybody else.".
This way, the four cooks have to talk to each other as little as possible, which is very good in software development -- not necessarily because they hate each other, but for other reasons like physical location, efficiency in making decisions etc.
It also means you can recombine the dishes (services) as you like. For example, you might just use the dessert to service a cafe, or just take the soup and combine it with a bread you bought from another company to provide a cheaper menu, or let other restaurants use your salads to combine with their dishes, etc.
One of the most successful implementation of SOA was at Amazon. Because of their design, they could re-package their whole infrastructure and sell it as Amazon Web Service.
*This is only one aspect of SOA.
SOA is an architectural style but also a vision on how heterogeneous application should be developped and integrated. The main purpose of SOA is to shift away from monolithic applications and have instead a set of reusable services that can be composed to build applications.
IMHO, SOA makes sense only at the enterprise-level, and means nothing for a single application.
In many enterprise, each department had its own set of enterprise applications which implied
Similar feature were implemented several times
Data (e.g. customer or employee data) need to be shared between
several applications
Applications were department-centric.
With SOA, the idea is to have reusable services be made available enterprise-wide, so that application can be built and composed out of them. The promise of SOA are
No need to reimplement similar features over and over (e.g.
provide a customer or employee service)
Facilitates integration of applications together and the access
to common data or features
Enterprise-centric development
effort.
The SOA vision requires an technological shift as well as an organizational shift. Whereas it solves some problem, it also introduces other, for instance security is much harder with SOA that with monolithic application. Therefore SOA is subject to discussion on whether it works or not.
This is the 1000ft view of SOA. It however doesn't stop here. There are other concepts complementing SOA such as business process orchestration (BPM), enterprise service bus (ESB), complex event processing (CEP), etc. They all tackle the problem of IT/business alignement, that is, how to have the IT be able to support the business effectively.
SOA is acronym for Service Oriented Architecture.
SOA is designing and writing software applications in such a way
that distinct software modules can be
integrated seamlessly with high degree
of re-usability.
Most of the people
restrict SOA as writing client/server
software-web-services. But it is too
small context of SOA. SOA is much
larger than that and over the past few
years web-services have been primary
medium of communcation which is
probably the reason why people think
of SOA as web-services in general
restricting the boundaries and meaning
of SOA.
You can think of writing a database-access module which is so independent that it can work on its own without any dependencies. This module can expose classes which can be used by any host-software that needs database access. There's no start-up configuration in host-application. Whatever is needed or required is communicated through classes exposes by database-access module. We can call these classes as services and consider the module as service-enabled.
Practicing SOA gives high degree of
re-usability by enforcing DRY [Don't
repeat your self] which results into
highly maintainable software.
Maintainability is the first thing any
software architecture thinks of - SOA
gives you that.
As far as I understand, the basic concept there is that you create small "services" that provide something useful to other systems and avoid building large systems that tend to do everything inside the system.
So you define a protocol which you will use for interaction (say, it might be SOAP web services) and let your "system-that-does-some-business-work" to interact with the small services to achieve your "big goal".
I would suggest you read articles by Thomas Erl and Roger Sessions, this will give you a firm handle on what SOA is all about. These are also good resources, look at the SOA explained for your boss one for a layman explanation
Building a SOA
SOA Design Pattern
Achieving integrity in a SOA
Why your SOA should be like a VW Beetle
SOA explained for your boss
WCF Service Performance
what tends to happen in large organizations is that over time everything is either monolithic or disparate systems everywhere or a little of both. Someone eventually comes in and says we've got a mess. Now, you want to re-design (money to someone) everything to be oriented in a sort of monotlithic depends on who you pay paradigm but at the same time be able to add pieces and parts independently of the master/monolith.
So you buy Oracle's SOA and Oracle becomes the boss of all your parts. All the other players coming in have to work with SOA via a service (web service or whatever it has.) The Oracle monolith takes care of everything (monolith is not meant derogatory). Oh yeah, you got ASP.NET MVC on the front or something else.
main thing is moving things in and out of they system without impact and keeping the vendor Oracle SOA, Microsoft WCF, as the brains of it all. everything's all oop/ood like, fluid, things moving in and out with little to no impact, even human services, not just computers.
To me it just means a bunch of web services (or whatever we call them in the future) with a good front end. And if you own the database just hit the database and stop worrying about buzzwords. it's okay.
Only one suggestion:-
Read SOA Concepts, Technology and Design by Thomas Erl.
It has very beautifully given the details about SOA in plain English and with case studies.
Well You see.. SOA stands for Service Oriented Architecture.... In simplest words, you write a piece of code that is very generic i.e. it does some thing that can be used in a lot of applications ... may be something like a address book or may be a calculator. and you launch this code on the IIS. So you provide a service through your code. So you are a service provider. Now someone wants to use a similar code then he does not have to write the code again. He simply uses your code maybe through a web service. Hence he becomes a service consumer. Hence making a program using such services is called SOA. And the loose coupling is there as the service provider and consumer may be interacting even if they are using diff programming languages.
Hope you understand.
from ittoolbox blogs.
The following outlines the similarities and differences to past design techniques:
• SOA versus Structured Programming
o Similarities: Most similar to subroutine calls where parameters are passed and the operation of the function is abstracted from the caller - e.g. CICS link and execute and the COBOL CALL reserved word. Copybooks are used to define data structure which is typically defined as an XML schema for services.
o Differences: SOA is loosely coupled implying changes to a service have less impact to the consumer (the "calling" program) and services are interoperable across languages and platforms.
• SOA versus OOA/OOD
o Similarities: Encapsulation, Abstraction and Defined Interfaces
o Differences: SOA is loosely coupled with no class hierarchy or inheritance, Low-level abstractions - class level versus business service
• SOA versus legacy Component Based Development (CBD) - e.g. CORBA, DCOM, EJB
o Similarities: Reuse through assembling components, Interfaces, Remote calls
o Differences: Wide adoption of standards, XML Schemas vs. Marshaled Objects, Service Orchestration, Designing for reuse is easier, services are business focused vs. IT focused, business services are course grained (broad in scope)
• SOA (for integration) versus Enterprise Application Integration (EAI)
o Similarities: Best practices (well defined interfaces, standardized schemas, event driven architecture), reusable interfaces, common schemas
o Differences: Standards, adoption, and improved tools
Reading the responses above, it sounds to me that SOA is what developers (good ones at least) have been doing from day one.
It could also stand for "Struct of Arrays" (as opposed to "Array of Structs") which is a common topic in parallel (especially SIMD) programming, but I'm guessing that's not what you mean here!
SOA is a buzzword that was invented by technology vendors to help sell their Enterprise Service Bus related technologies. The idea is that you make your little island applications in the enterprise (eg: accounting system, stock control system, etc) all expose services, so that they can be orchestrated flexibly into 'applications', or rather become parts of aggregate enterprise scoped business logic.
Basically a load of old bollocks that nearly never works, because it misses the point that the reasons why technology is the way it is in an organisation is down to culture, evolution, history of the firm, and the lock in is so high that any attempt to restructure the technology is bound to fail.
Have a listen to this week's edition of the Floss Weekly podcast, which covers SOA. The descriptions are pretty high level and don't delve into too many technical details (although more concrete and recognizable examples of SOA projects would have been helpful.
A traditional application architecture is:
A user interface
Undefined stuff (implementation) that's encapsulated/hidden behind the user interface
If you want to access the data programmatically, you might need to resort to screen-scraping.
SOA seems to me to be an architecture which focus on exposing machine-readable data and/or APIs, instead of on exposing UIs.
SOA or Service-Oriented Architecture is a software architecture pattern in which applications or systems are constructed from underlying (and usually distributed) software services that conform to a specific set of characteristics, namely:
Interface, Policy and Contract based
Location transparency
Autonomous
Abstract
Reusable
Composable
Stateless
Discoverable
Extensible
Loosely coupled
The primary goal of SOA is sofware development agility, i.e. the ability to respond the change easily, and cheaply, thus allowing businesses to rapidly respond to changing markets.
Services are typically (but by no means exclusively) implemented as web services, i.e. they operate over the ubiquitous web HTTP protocol, and are implemented either using XML-based SOAP or the lightweight (and more popular) REST paradigm.
Depends on who you are!
If you're an business owner, SOA is a solution to increase your incomes and business agility. If you're an entreprise architect, SOA is a way to draw nice and clean piece of software on a blank canvas. If you're an architect SOA is the solution to design loosely coupled services over an integration platform, to just plug services into outlets. If you're a developper SOA is a programming paradigm where a service is in the center of the design and the code.
You should read 100-SOA-Questions [pdf]
Cheers
Service Oriented Architecture (SOA) is a software architectural style that builds applications as a collection of pluggable parts, each of which can be reused by other applications.
Maybe I just expected "three-tier architecture" to deliver a little more than just a clean separation of responsibilities in the source code (see here)...
My expectations to such a beast that can safely call its self "three-tier architecture" are a lot higher... so, here they are:
If you were to build something like a "three tier architecture" system but this time with these, additional requirements and constraints:
Up and running at all times from a Users point of viewExpect when the UI gets replacedWhen other parts of the system are down, the UI has to handle that
Never get into a undefined state or one from which the system cannot recover automatically
The system has to be "pausable"
The middle-tier has to contain all the business logic
Obviously using an underlying Database, itself in the data-tier (if you like)
The business logic can use a big array of core services (here in the data-tier, not directly accessible by the UI, only through business logic tier facade)
Can be unavailable at times
Can be available as many parallel running, identical processes
The UI's may not contain any state other than the session in case of web UI's and possibly transient view baking models
Presentation-tier, logic-tier and data/core-services-tier have to be scalable independently
The only thing you can take for granted is the network
Note: The mentioned "core services" are heavy-weight components that access various external systems within the enterprise. An example would be the connection to an Active Directory or to a "stock market ticker"...
1. How would you do it?
If you don't have an answer right now, maybe read on and let me know what you think about this:
Sync considered harmful. Ties your system together in a bad way (Think: "weakest link"). Thread blocked while waiting for timeout. Not easy to recover from.
Use asynchronous messaging for all inter-process communication (between all tiers). Allows to suspend the system anytime you like. When part of the system is down, no timeout happens.
Have central routing component where all requests get routed through and core services can register themselves.
Add heartbeat component that can e.g. inform the UI that a component is not currently available.
State is a necessary evil: Allow no state other than in the business logic tier. This way the beast becomes manageable. While the core services might well need to access data themselves, all that data should be fed in by the calling middle tier. This way the core services can be implemented in a fire and forget fashion.
2. What do you think about this "solution"?
I think that, in the real world, high-availability systems are implemented using fail-over: for example, it isn't that the UI can continue to work without the business layer, instead it's that if the business layer becomes unavailable then the UI fails over to using a backup instance of the business layer.
Apart from that, they might operate using store-and-forward: e.g. a mail system might store a piece of mail, and retransmit it periodically, if it can't deliver it immediately.
Yep its the way most large websites do it. Look at nosql databases, Google's bigtable architecture etc.
1. This is the general approach I'd take.
I'd use a mixture of memcached , a nosql-cloud (couch-db or mongo-db) and enterprise grade RDBMS systems (core data storage) for the data layer. I'd then write the service layer ontop of the data layer. nosql database API's are massively parallel (look at couchdb with its ngingx service layer parallizer). I'd then provide "oldschool each request is a web-page" generating web-servers and also direct access to the service layer for new style AJAX application; both these would depend on the service layer.
p.s. the RDBMS is an important component here, it holds the authoritative copy of the all the data in the memchached/nosql cloud. I would use an enterprise grade RDBMS to do data-centre to data-centre replication. I don't know how the big boys do their cloud based site replication, it would scare me if they did data-cloud to data-cloud replication :P
Some points:
yYu do not need heartbeat, with nosql
the approach taken is that if content
becomes unavailable, you regenerate it
onto another server using the
authoratitve copy of the data.
The burden of state-less web-design
is carried to the nosql and memcached
layer which is infinitely scalable.
So you do not need to worry about
this. Just have a good network
infrastructure.
In terms of sync, when you are
talking to the RDBMS you can expect
acceptable synchronous response
times. Your cloud you should treat as
an asynchronous resource, you will
get help from the API's that
interface with your cloud so you
don't even have to think about this.
Advice I can give about networking
and redundancy is this: do not go for
fancy Ethernet bonding, as its not worth
it -- things always go wrong. Just
set up redundant switches, ethernet cards
and have multiple routes to all your
machines. You can use OpenBSD and
CARP for your routers, as they work
great - routers are your worst point of failure -- openbsd solves this problem.
2. You've described the general components of a web 2.0 farm, so no comment:D