← Previous lessonFree lesson 3 of 3 · lesson 3 of 48

Unlock all 48 lessons of Microservices using Java, Spring & DockerFull course ₹40,000 · paid

Price & syllabus
Free lesson · 20 min · no payment

Comparing Architectures and Introducing Microservices Fundamentals

A free lesson from Skill-Lync's Microservices using Java, Spring & Docker course, on Grow by Skill-Lync · Last updated 16 Nov 2026 · Session: “Week 1 Session 2”

What will you learn in this free Microservices using Java, Spring & Docker lesson?

The instructor compares monolithic, SOA and microservice architectures on implementation and maintenance, scalability, testability and fit for agile development, using Uber and Flipkart examples. The lesson then introduces microservices through a bank loan service and explains cloud native applications and cloud auto scaling.

  • Implementation and maintenance trade-offs
  • Scalability and testability by architecture
  • Waterfall versus agile development
  • Cloud native applications and auto scaling

What comes after lesson 3?

Later in Skill-Lync's Microservices using Java, Spring & Docker course (paid): lesson 4, Week 1 Session 3; lesson 5, Week 1 Session 4; lesson 6, Week 2 Session 1. The full course has 48 lessons (18 h 7 min of video); see the course overview or its price and syllabus.

Lesson transcript

Machine transcript of the lesson audio (lightly cleaned). Technical terms may contain errors; the video is the reference.

Hello everyone and welcome back to the second session on microservices. In this session, now we'll continue with our discussion of the comparison of various architectures that we have seen so far. So as we saw, there are three types of architectures. One was tightly coupled, the second was loosely coupled and third was decoupled. So we also saw like you know how they are architected and what are the various parameters by which we designed them. So now let's try to understand what are like you know how they compare with one another. So the first attribute which we'll use for comparison is the implementation and maintenance. That is one of the key parameters which we use to evaluate any architecture. That is how easy it is for us to kind of implement that architecture and also maintain it over the time. So as we saw that for Monolithic, the implementation and maintenance is easiest among the others architecture. Why that is the case? Because since we are keeping all the presentation layer logic, the application service layer logic, data access layer logic, everything as in one code base and also deploying it to a single server, it kind of becomes easy for us to maintain that application and also develop it easily. But if you take any architecture like which is loosely coupled like SOA or event driven architecture, the implementation and the maintenance becomes partially complex because now we have to deal with developing independent systems which are deployed on different servers and also like you know, we have to also find a way to interface between each other. So that is why the implementation for loosely coupled server is partially complex.

And finally, if you take the decoupled architecture like microservices, the implementation and maintenance is complex and all these are kind of relative when I say the easiest and when I say complex. So these are just related to one another. So if I have to just rate implementation and maintenance between all the three architectures, then I would say the microservice architecture is most complex to implement and monolithic architecture is the easiest to implement. Right. And why we say it is complex for microservice architecture. That is because as we saw that you know, we have to develop independent services and database for each of the services, which will be again individually independently deployed on different servers. So in this case, the kind of code and overhead which we have to write and manage is large. That is why the implementation and maintenance for microservices is complex. The next attribute which will be used for comparing is scalability.

scalability. So what we mean by scalability is so I'll give you a very simple example. So assume that you know, you are developing an application for booking cab. Okay, so you just imagine that you are a startup who just invented a cab booking application. So as you know, when you just initially develop your application, there won't be many users, right? Probably it will be just 10 users or 100 users. But just assume that you know, you have you are started to grow like Uber. So over the time you have really garnered like you know, tens and thousands of customers. So what happens is over the time as your application popularity grows, even the load of your server on your server improves, it goes up, right?

So today there will be 10 users accessing the application tomorrow, there will be 100, then day after there will be 1000, so on and so forth. So when we say scalability, all it means is does our application grows over the time as the load on the server increases? Okay, so in case of monolithic application, which is like in case of the tightly coupled architectures, it is not easily scalable. Why it is not easily scalable is since we are going to kind of just deploy all the application in a single server. So what happens is like, you know, whenever there you have to scale your application or your application requirement increases, either you have to make changes and no redeploy all your application in the server, or you have to kind of add an extra server manually and then deploy all this application into it. So the kind of like, you know, the work which you have to do to make it scalable is not easy. Okay, and mostly monolithic application are also on premise server. What do you mean by on premises, since you are going to have a server physically in your environment where you're going to deploy it. So this maintenance of server and scaling of the server is not quite an easy job.

Because you cannot just add another server on the go without first having them in your environment for loosely coupled architectures like Sova, if the scalability is partially complex, why? Because like, you know, since it is loosely coupled, the components can be independently scaled, but still there are certain limitations, which has to be taken into consideration. They are just not like, you know, infinitely scalable, you also have other complex parameters in place, because they are going to be interconnected using service bus and you know, there could be other server related restrictions.

functions as well. But if you take architecture decoupled architecture like microservices, scalability can be easily achieved. Why that is the case? Because since your servers are independently deployed in the cloud, and as you people may know in the cloud, it is very easy to spin up a new instance and create a new instance of your server just on the go, right? You don't really need to have all this overhead of like, you know, manually provisioning a server, then starting the server, deploying your application and all that.

So when you use any of these cloud providers, like AWS, Google, or like Azure, all of them provide this auto scaling option. So that once you were in once you deploy your microservice application, it is very easy for us to kind of scale our application on the go. So what happens is tomorrow if you as I said, if you say develop a app, which is a Uber competitor, and over the time your application popularity increases. So you don't really need to sweat a lot in order to make your application handle the large load.

If you enable auto scaling and other such features which are available with the cloud vendors, your application can easily grow over the time. The next important attribute for our application is testability. So what we mean by testability is how easy is it for us to test our application, right? For monolithic, the testability is really not scalable or not easy. Why that is the case? Because since everything is kind of tightly coupled, it is not easy for us to independently kind of test each and every part of the application.

independent of each other. So either you have to write a test case, which will test end to end scenarios, but there you cannot test a few scenarios where you just want to slice and dice your application and test just a few functionality of it. So say for example, like, you know, if I develop a application for IRCTC, right, for example, and I just wanted to test the part where like, you know, I want to test if the logic for Tatkal is working fine. If it is a monolithic application, I have to write the test right from the user kind of like, you know, clicks on the booking reservation system, right?

Right. Till the point that the ticket is generated because my entire code is so tightly coupled. I cannot just take out a single functionality and test it independently, right? But whereas in case of loosely coupled architectures like so what testing is partially doable or I would say like, you know, we can achieve a good sense of testing test cases. Why that is the case? Because since we are going to develop loosely coupled components, it becomes easy for us to kind of isolate the few components and test each test them independent of the other components.

But whereas in microservices, since they are totally independent of each other, the testing can be easily achieved and you can easily slice and dice your application to write your test cases accordingly at the last parameter, which we are going to discuss today for this comparison is fit for agile development. So first we should understand what this agile development, right? So the people, some of you are not familiar with this agile development concept. So in the last, I would say like 10, 15 years, this, there is this concept of agile development, which is now the industry standard for developing any application.

So earlier, what used to happen is we used to follow the waterfall software development lifecycle, where in like, you know, each phase of your application had a definite step. And, you know, we used to develop and complete the application in its whole. So say, for example, if I had to write an application for Uber, right, it was not like, you know, I could just write a simple prototype this week, test it and then ask my user how it is. And then again, keep on improving it. How would the Uber used to be developed in those days? What I mean to say is like before agile came into picture is first used to take the requirement from the end user from the business users, then you kind of analyze those requirement.

Then you kind of like, you know, wrote the code for that particular requirement, then you tested it, then it went for user acceptance testing. And after like every all the approvals was received. And then finally, what the software made made available to the end user, right? So this used to be really taking a lot of time, sometimes even years, and it was really cumbersome process. So that's why what happened is now it has been replaced with agile development. So what agile methodology says is instead of like, you know, waiting for like, you know, quite a lot of time and doing a lot of redundant work on that aspect, it is better that you know, you just keep on iterating incrementally and develop value adds just over the go.

So say for example, if I'm going to develop a Uber app, maybe in the next one month, I will develop a simple app, which is just going to let me book a cap from say point A to point B. That's it, no other features enabled. So maybe the next iteration of that probably I would like, you know, even get a feature where I could schedule a ride, maybe some days or hours in advance, maybe in the third iteration or the third release, I would have a feature where you know, maybe I could book for my friend or relative, or maybe and in the next iteration, maybe I could have something like a profile where I could save my card, you're getting my point, right?

So the whole point is if the same was done in waterfall, what used to happen is all these features usually used to be developed, used to be given or released in one single build. And that used to really take a lot of time because like, you know, developing all this together used to really be a time consuming process. But in agile, this is like, you know, we have kind of made things loosely coupled and we the focus is more on the value ads or the end user, we are doing incremental deliveries threat. So why we were talking about agile is since agile is I said the de facto standard for the today's software development lifecycle tightly coupled or like, you know, architecture like monolithic are not easily fit for agile development.

Because as I said, monolithic architecture are really a good fit for the water cycle kind of lifecycle of development because there you They are partially developed a lot, like, you know, the entire application in one go and deployed. And if you take loosely coupled architectures like SOA, they are partially a fit for agile development. Why I say partially? Because since they have components which can be like, you know, independently developed, they can be developed by various team in parallel. But there are again a few bits and pieces which needs to be have, which has some dependencies with all these components. As I said, like, you know, you have these service bus and other infrastructure related setup. So that's why it's not a full fit for agile development.

Whereas if you take any decoupled architecture like microservices, it is a perfect fit for agile. And why that is the case? Because each services, since they are independent and like, you know, you can just develop them independently, you can have multiple teams which can work in parallel. So for the same Uber application, what I can do is one team can work on the feature where they can book the cab. The other team can book on a feature where they can like, you know, book the cab in advance. The third team can work on a feature where maybe they can book the cab for friends or relatives.

The fourth team can work on profile, so on and so forth. So that is why decoupled architecture is the best fit for agile development. So moving ahead is now that we have understood what are the various architectures available. So, so far, whatever architectures we saw, these were just to ensure that, you know, you have a broad idea of what we are doing. Right. So before just jumping into microservices, it is also important for you to know what are the other alternatives available in industry today and why we are learning microservice.

So I hope by now you would have understood that microservices has a lot of plus points compared to other architectures, mainly like, you know, in terms of what we saw right now, right? It is easy to implement. I was sorry, I won't say easy to implement. It is like, you know, easy to scale, easy to test. And also, it's very a perfect fit for agile development. We also saw the challenge that, you know, it is a bit complex to implement and maintain. Right. But in this course, we'll also see how we can mitigate that to certain extent, right? So without wasting much time, let's now jump into the introduction of microservices.

I'll just dive a little deeper into microservice architecture and understand what is all about. So microservice is a pattern in which the application with several independent services is composed. So as you can see on the picture on the right hand side. So as we have a user interface, which we'll be talking with the API gateway, API gateway is nothing, but you can just say it is a component which will just listen to the request, which is called the API gateway. It's coming from the UI and then decide to which microservice this request has to be routed to. Okay.

So what happens is when you develop an enterprise application, normally the user interface will not directly tell the server which microservice or which server this request has to go. All they will have is they will have a certain request in a format and they will have certain like, you know, payload attached to it. Then your API gateway will dynamically decide then, okay, which request, which microservice server will this request go to? Does it goes to the HR microservice or the like, you know, I'm just giving a simple example from my enterprise.

So does it go to HR and microservice or does it go to the operation microservice? Does it go to the series microservice? So on and so forth. So microservice is just kind of like, you know, the API gateway is nothing but just kind of an interface which sits between the end client and the microservices. And each of these services that you see below, right, microservice one, two, three and all that. So all each of these service specializes in performing one particular task independent of other service.

Okay. So say for example, let's take an example of bank which we are going to develop in this course. So if I take a bank and like in a bank as many departments as you know, so in that if I just take the example of say loan department. So one specific function of loan department is to give loans, right? So I will have a API or I will have a microservice which is designed only to grant loans. Okay. So that API's only function is to ensure that based on the request received, I will either approve a loan or create a give a loan or I will reject that loan.

Right. I won't do any other additional task with that. For example, I won't give you the balance or like, you know, I won't tell you what is the status of your account and all for that. I will have to develop a different microservice. So the moral of the story is microservice. Each microservice will only specialize in performing a particular task and that has to be independent of other service. That is a key point to be noted over here. And since microservice applications are composed of several independent services, they are also easy to develop, test and deploy independent of other services.

Okay. So don't get confused. In the earlier comparison, we said microservices are generally little complex to develop and maintain. Right. But here we are saying they're easy to develop, test and deploy. So let's understand what it means. So here. So as I said, today we have agile development methodology. Right. Where the focus is on like, you know, creating small, small features incrementally over the time and develop the entire application. So what happens in the enterprises normally for a particular application, there'll be many teams that should be working. Right.

And since each of these services are independent, it is easy for us to kind of split these each of these services into features and give it to them independent teams. So say, for example, if I have to develop Uber application or IRCTC application, so I can have, say, maybe five teams. One team can just focus on the train ticket reservation part. The second team can just focus only on the food ordering part. The third team can maybe just focus on maybe like, you know, the part where like you can also allow the user to book flight tickets, so on and so forth.

So what happens is all these three teams don't have to depend on each other. They can just simply start the development testing and deployment activities independent of each other. And finally, when all of them are complete, I can just integrate all of these into my core application and we are good to go. So that's why we say that, you know, since they are independent services, they are easy to develop, deploy and test. Okay, and why they are also easy to test because since they are independent, every team can write their own test cases.

They don't really need to depend on other team to write their test cases and also micro services are distributed and decoupled from one another. So they can be scaled very easily. So what is the meaning of that is since we are kind of like, you know, decoupling each and each service from one another. See, this is a term which we are repeating again and again, right? We are kind of we are seeing that, you know, it is an independent service. It is decoupled service. So what what is the benefit of that is they can be scaled very easily.

Okay, so I'll give you an example of this. So just assume that, you know, we have developed maybe like, you know, we have developed an e-commerce website. Okay, so maybe just take the example of Amazon or Flipkart. So what happens is Amazon or Flipkart are really made of tons of microservices, right? So assume that, you know, some festive season comes in, maybe like, you know, Diwali, Holy Eid, whatever. So suddenly there is a spike in the people searching for a particular product. Okay. So what happens is we want to ensure that our application is able to handle that search query faster.

So maybe the people are loading your application and searching for, say, maybe a kurta or pajama, whatever. So your microservice, which is concerned with that particular search feature, right? Which is searching for the user query should be scalable at that particular period of time. So in case of microservice, it is quite easy to do that because since our services are independent, we can spin multiple versions of multiple, like, you know, deployments of that particular service into other services. As I said, cloud vendors support this auto scaling features.

So we don't really have to sweat a lot and write a tons of code or do a lot of configuration in order to achieve that. Similarly, if you have to develop cloud data applications, microservice architecture is one of the necessary ingredient. Okay. So what is the meaning of cloud native application? Don't worry. We have a in-depth discussion about this in the next week session. But for now, just understand that cloud native applications as a name implies are the applications which are written for cloud. Okay. So what is cloud? If you people are not aware, we basically have like, you know, there are, there are three famous vendors of cloud.

Like one is Azure, Google and AWS, Amazon Web Service. So what cloud means is earlier, whenever you used to write applications, we used to kind of like, you know, have the service for that particular applications is like, you know, installed or bought out. And, you know, we used to maintain our own server farms, right? That enterprise, you really used to maintain our own server farms. But over the time they realized like, you know, it is really a waste of effort and energy and all that thing to like, you know, have all that ourselves and, you know, maintain it.

Because say, for example, if you take a company like Uber, right, their core strength or their core revenue is from like, you know, writing the code, which helps you to get the cab and, you know, do all that stuff. But if you're going to spend a lot of energy just to maintain the server, deploy the server, monitor the server, like, you know, they really are spending part of the energy which should be used productively in some other part of the application. So that is why this cloud came into picture, where we are outsourcing the infrastructure part of our application to some third party vendor like, like, you know, Microsoft, Google or Amazon, wherein they have the infrastructure upon which we can deploy our application.

That is just a very simple definition of cloud. Cloud also has other advanced features, wherein you can also have certain services also has certain services outsourced. So say, for example, you want to like, you know, do SMS or like, you know, you want to send automated emails. So even those kind of services, instead of writing your own application, you can simply subscribe to the cloud vendors and get that service provided from them. Okay, so and also another thing to notice, the industry is now slowly moving into the cloud native cloud native or cloud first application.

So wherein any application that you develop has to support cloud as well. And in order to develop cloud native application, microservice architecture is one of the essential ingredient. Why? Because since microservice is develops like kind of independent services, it is one of the necessary ingredient in order to ensure that, you know, it is able to run in the cloud and able to scale fast. And with this, we have come to the end of this session. In the next session, we will now next understand what is the Java ecosystem for microservices.

Thank you.

Want the rest of Microservices using Java, Spring & Docker?

Get the full course syllabus and fees.

The full course is paid (₹40,000). We'll send the syllabus and call you back within one working day.

Talk to a career counsellor

Free 20-minute call. We'll map the right course to where you want to work.

By submitting you agree to our Terms and Privacy Policy.

Full course · paid

Unlock the full Microservices using Java, Spring & Docker course

The remaining 45 lessons, mentor-reviewed projects and a certificate come with the paid course (₹40,000). Leave your details: we'll send the syllabus and fees and call you back.

Get the syllabus and fees

A counsellor calls you back within one working day.

By submitting you agree to our Terms and Privacy Policy.

Grow by Skill-Lync · free lesson

Start watching

Free, no payment. Leave your details once and every free lesson on Grow plays straight away. A Skill-Lync counsellor may call you about the full Microservices using Java, Spring & Docker course.

By submitting you agree to our Terms and Privacy Policy.

Talk to a career counsellor

Talk to a career counsellor

Free 20-minute call. We'll map the right course to where you want to work.

By submitting you agree to our Terms and Privacy Policy.