Jakarta EE Tech Talks, 27 Apr 2020

Previous Meeting Next Meeting

⏯

youtube image

►

From YouTube: Build a Microservice from scratch with Microprofile (Helidon) | Jakarta Tech Talks - February 2020

Description

Microprofile is one of the key topics in Java development right now. Everyone is talking about Microservices and cloud-native development. This talk is all about how you can start developing a microservice with Microprofile specifications. This talk will have lots of code examples with Helidon framework and developers will have a deep dive into the Microprofile concepts.

///GET SOCIAL!

Subscribe: https://www.youtube.com/user/EclipseFdn
Follow us on Twitter: https://twitter.com/JakartaEE
Like us on Facebook: https://www.facebook.com/JakartaEE/
Join us on LinkedIn: https://www.linkedin.com/groups/13597
Tag us: #JakartaEE
Want more great Eclipse Community webinars? Become a member of our Jakarta Tech Talks Meetup: https://www.meetup.com/jakartatechtalks

A

Hi, everyone welcome back to another to credit, Tech Talk I'm, your host, Laura and I'm here to introduce our guest speaker today, Tom, who will be presenting about micro profile and hello Don. If you have any questions, please leave them and ask a question button below or you can ask them in the chat box over to you. Tom thanks, Laura, hello, everyone! So today, I'm going to talk about how you can build micro service using the micro profile and the hilly done framework, so I am Sam good morning.

A

So before we get started little background about me, so I'm, a Java developer and I'm working as a full time employee in software, AG I'm, an API evangelist, I talk about API and API based practices and the API programs a lot. Apart from my regular work, I'm, a Madras Java user group bleed and I'm married and having two kids and when my free time, I, usually blog on technologies.

A

Recently I am fascinated about the garden engineering, so trying different things with gardens and all the things.

A

So what we are going to discuss today, I hope. You all know this. Some understanding about the micro services, so I'm not going to cover in detail about micro services, but definitely I will cover what is micro profile and why micro profile is so important and then I will move on to a Kelly down framework, which is the micro profile compliance framework and then I will give you a walkthrough on each and every specification driven in the micro profile.

A

First, we we'll start with what is micro, profiler, so micro profile is a programming model to develop micro services in the enterprise Java world, so it is completely driven by the community and there are lots of good things it taken from the Jakarta EE framework and it is built on top of it, and it also offers lots of cloud native development practices, something like the health check and fault tolerance. All these things will come into the picture, so we'll go deeper and there are lots of active contributors behind this micro profile initiative.

A

So just to name few, we have IBM red, add Tomi, tribe, payara, Oracle, Microsoft and many more so why micro profile is so important. Now.

A

If you see few years back the enterprise Java development, there seems to be a complaint that it is heavy and it is very hard to build application and it bundled some some most of the unused libraries. There are lots of hiccups here and there. So the main agenda behind micro profile is eliminate all these burden. When you develop the micro services and you don't need to learn anything new apart from your existing learnings.

A

So if you know Java and the Jakarta yi learnings, then you can take those land learnings and you can get started with the micro profile development and you can choose your vendors meaning.

A

There is no lock-in with the micro profile implementation if you code or if you develop your micro services based on the specification and then you are free to move between different implementation vanguards.

A

So there is no lock-in with the vendors and there are multiple ways of using their api's, especially they offers the CDA support, and you can also interact with the treu program and they are also offering CLI and it is completely specification driven meaning every I mean the decision is made through the community process and all these things will be reviewed by the community and it is completely backed up by the community. So everyone is free to give their opinion and you can also improve all the micro profile standards and everything.

A

So that is the beauty behind the whole ecosystem.

A

Okay, we have seen why micro profile is important and what is micro profile now we will go little deeper and the current version of micro profile. What is all about micro profile so underlying is the simple and well-known jagat ie a good part. So if you, if you see in the four boxes, CDA everybody in all CDI and jax-rs, Jason, B and JSONP, all these are fundamental to build any web services in the Java ecosystem. So they have taken these core pieces as a fundamental component and top of it.

A

They also offer lots of cloud native specification on top of it, especially the configuration fault, tolerance, scale, check, Jade ability, earth matrix and open API tracing and risk land. So today, I am NOT going to cover the Jade ability earth other than that all the other parts I will be covering so on top of it. So the complete annotation support is available for your development, just to ease or simplify your development lifecycle.

A

So there are three things is marked in the orange color. There is a specific semantic behind that these three specification is modified or updated from the earlier micro profile. Specific, that's why I have marked these three things alone in a different color. Well, the remaining things. There is no change okay, so we have seen the micro profile specification now who is supporting or who is offering a different kinds of framework to build our micro services.

A

So, as we seen earlier, there are a plenty of opportunity we have in the market right now, so we have Parker's developed by Red Hat and we have open Liberty supported by IBM, and we also have a lead on. Of course, today, I'm going to cover Kelley done, Sohail Eden is being developed by Oracle, so likewise we have many different vendors. They also offer different kinds of framework. All these frameworks are completely complaints with micro profile.

A

Okay, now we move on to the important topic Hayley done and let me spend a few minutes to explain for what is really done. So harridan is all about a micro profile, implementation framework. It is developed by Oracle and it is completely supporting docker and kubernetes from from day one meaning. Nowadays, everybody is going with the containerized deployment model when it comes to the micro services, so they are supporting this from day one. So they, when you create the skeleton code, they also offer.

A

These kind of I mean skeleton code along with the generated codes, and it is completely powered by Nettie one of the reactive webserver available in the market, and they have two different various variants in the micro profile. One is the micro profile variant and another one is the standard edition and they also support the grail vm engine. So if you generate the native binaries, you can directly say deploy into the growl vm engine.

A

Okay, now we will dive deeper into the how we can I mean get started with the real framework. So this is simple, maven archetype, so you can customize it based on your business needs and you can change the package structure and the group ID and everything. So once you execute this, you will get it get this really skeleton code which will be downloaded into your system and from that you can get started. If you are not comfortable with the executing the this archetype, then they also offer a different UI, so this one was launched recently.

A

So if you visit start dot micro profile, dot, ivo, you will get, I mean the nice UI and there you can fill the necessary information and the moment you keep the download button, your binary will be downloaded, meaning your skeleton code will be downloaded and you can get started from that. So, basically, in the, if you see in the right side you, there are a few options this mentioned, so you can pick and choose which library you want to include in your micro services development.

A

If you don't need the open, API or the open tracing, then you can just uncheck and only those dependency will be bundled in your final wineries.

A

Okay, so I'll get started with this simple Jack service web services. Everybody is talking about micro services, so we we also build some micro services from that. So I am I, have already created a skeleton code by executing the maven our archetype. So this is already imported into my IntelliJ editor, so you can use any editor based on your comfort, so I'm using IntelliJ and uncomfortably that. So this is the basic scale. Ten code is generated and if you see the main file, so they are generating a simple server and they are starting it.

A

So this is a starting point and once the skeleton code is ready, we'll check whether this compilable or not and I'll, just execute this maven command.

A

A

Yes, it is compiled. My build is successful. Then you can start the server using Java jar and with the final output of my halidom jar. So this will internally start my server with the port 8080.

A

Okay, my server is planted so as I mentioned earlier, so we'll get started with building some restful web services and how we can handle exceptions in the web services and how you can generate some links and a few example with that. So for that, I will open a simple service to explain this concept, so the interface so I usually start with the interface and then do all the annotation in the interface and then my implementation code will be very clean.

A

So this is how you can create a simple web services, and here it's a pink service which has two methods. One is pink and other one is the welcome message and if you see implementation, there is nothing I'm. Just returning some messages in this in this I'm, just returning a hot coded message.

B

And in the Welcome message, I'm just taking something from the configuration so I will.

A

Take I will cover this in the next okay, so let me check whether this is working. Fine are not so I'm just checking my postman, and this is my pink service.

A

Just clicking yep, it is working fine, so we have seen how we can create a simple web services using jax-rs specification and next we will see how we can handle the exception. Part so far the exception. So it's for every business we may have yep they're. Just asking Canon watch your text a little bit. Okay, it's like yeah.

B

A

A

Just a minute, I'll increase the font size.

A

A

Hope this would be clear now, okay, so this is my business exception. So whenever we build some application, we need to build some custom exceptions and how these custom exceptions could be handled by the jack soloff's specification itself. So.

B

This is the exception mapper and if you see here so the this exception, mapper is bind with my custom exception, and this is annotated with provider as Phyllis the produces JSON media type. And if you see here this is how I'm generating the response, whenever my application throws some exception and automatically this this exception, mapper will handle.

A

B

My exception so I don't need to generate the appropriate response based on this exception, so it's all I'll simply throw this exception from my code. So.

A

If you see here.

B

A

This is the country API and if you see here so this will be thrown. So if you go to the postman and I'm just getting all countries.

A

Yeah, so I I will explain how I'm handling the exception part. So if you see here this, this.

B

Exception response is being thrown from my exception as a response handler. So if you go to B.

A

This exception mapper, so basically my country API. If something is.

B

Not I'm having a manager and this manager is not able to identify based on the code, so I'll throw an exception. So either this code is null or the code is not able to identify a country I'll throw an exception, so this exception will be automatically identified by this exception, mapper and it will be transferred to the client and the next one is usually. Whenever we create some object, we will send the link park just to maintain our just to give the indication to the user how you can navigate to the appropriate resources.

A

Once you got the successful response, so this can be easily achieved using the link API. So we have a simple link generator here.

B

And this lead generator will get the path from the URI info.

A

And it will append the slash as.

B

Well, as it will add the Silph reference and based on that, it will be injected into the response header and from that header you can get the response back and you can process accordingly whenever you want to try a generate some link for a created resources.

B

A

Okay, so I have covered the jax-rs part and I will move on to the configuration part.

B

So this is one of the important and very interesting topic when it comes to the micro services development. So if you remember that 12:12 factors for building an app, that is a best practices, is being advocated by many industry. So if you see that they are advocating, one important component here is the configuration meaning. You should be able to transfer our paths at configuration externally.

A

Outer out I mean from the external field, so.

B

As part of the micro profile, so they have defined three layers of configurations where they are referring that like config source and it is started with the micro profile, configuration and the environment configuration and the system level configurations. So if you closely watch the slide in the right side, so I have put some number within.

A

This bracket so.

B

Meaning this is the ordinal number ordinal factor so based on this number, the priority will be defined. If you see here, the micro profile, ordinal number is true hundred and the system ordinal.

A

Number is 400, meaning.

B

Whenever you pass some variables at the runtime, so that will take the highest precedence than the micro profile variables, so this configuration is available as part of the resources folder.

A

B

A

The application level configurations you can add here and you can bundle it along with your application, so.

B

I have few configuration, it is a simple property file, and so you can add any any number of properties and you can load when the system starts. So I have mentioned a little server port and the host and some of the properties.

A

So how I can access.

B

These configuration properties inside my code, so a micro profile offers two different variants. One is the config property and the other one is the config annotation. So the conflict property is mainly used to access individual properties based on the key. So if you see this example, so AB dot greeting is a key available in the micro profile, a property file and which is accessed through config property annotation, but in in some cases we may need more than one property properties at the same class file.

B

So to address that use case, they have introduced something like config annotation so through which you can lo. Ok, you can get all the configuration property reference at one shot, so I'll just show the court how you can achieve that. So if you open.

A

This pink API, so I have injected. So all these.

B

Injection and everything is supported through CDI, so you can just inject this config reference and it will be automatically injected into the core and from this config and I can call the gate value.

A

With the AB dot welcome and you can pass the.

B

The value type so, in my case it's a string, but you can also do integer.

A

Long or whatever it may be, based on your requirement so.

B

This this one can be accessed from a welcome, so I will just change this thing.

A

So see, I have not passed any variable here.

B

So in this, in this examples, this conflict property expects some kind of variable here, so the variable has to be passed with this query parameter. So if I pass, this query parameter.

A

So this will be automatically injected into this variable.

B

And we will get the final passed string and then now I will show you how we, how we can I mean change some of the in a bundled property from the Journal world. So for that, I will take this key.

B

So I'll start the server and.

A

Before I start this server with the minus D flag.

A

So it is currently showing welcome celadon, so I I will change this to.

B

Hello, okay, so I will start the server now I am passing some variable through runtime, so you can also define your own custom config sources and it can be attached to your application. So you can also read the config file from the datastore database or some static files. So those things can.

A

Be handled at the micro profile level, so I, my server is started now and if I hit this.

B

It should return hallo.

A

Halidom, that's.

B

What my expectation- yes, it is working fine, so this is how we can pass some argument to change the configuration which is available at the micro services level, so the this is where this ordinal factor will come into the picture and it will play a role so sometime. If you bundle more than one configuration file, you can also enforce the cardinal factor for each of your property file say: for example, there is one more property available in the configuration specification so which is config underscore ordinal.

B

So if you add this property in to your configuration file, then that will also be treated when this. This configuration specification loads, the configuration property file into the application so based on this, it will overwrite if it finds more than one key across different property file. So the merge will happen based on merge are the over. It will happen based on the ordinal factor.

B

That's how the specification mentioned that okay, so I hope you have some understanding about the configuration and how you can pass some external configuration and how you can write the kinetic sources and everything. So next we will move on to the health check. This is again an important component when it comes to the micro services, so each micro services should somehow indicate to the external application or the developers that how my help is so based on that, if the application Warner want to take some decision, the business I mean the business availability decision.

B

Then he can span a new instance or if he can also get notification. All these the DevOps cycle can be managed in a smoother way. So for that the health check specification offers slash. Health is the endpoint. You don't need to write anything to get that out of the box. It is, it is by default. They are supporting it, meaning every micro profile implementation vendor has to offer the slash health API so that this by default. This offers the kind of intelligence one. Is the deadlock, detection and available a disk space and available heap memory.

B

So, based on this, the JVM level intelligence, we can take some important decision and span our extend our aliveness of our application. So I'll just go through how we can check.

A

B

So this is the end which is available by default, so we don't need to provide anything so the this. This will give you some indication about the disk space just pay attention to this point, the disk space and heap memory and.

A

And all the other important.

B

The deadlock yeah the deadlock, so these are the three things will be available out of the box. So apart from that, if you want to check the liveness and the readiness, so these are, the two changes was introduced in the recent specification of Cal check. So if you want to check the liveness as well as readiness on your own, so you can provide your own implementation on top of the existing API and you can get the benefit by this health check API. So.

A

B

Will see how we can get this done so they have introduced two kinds of annotation one is liveness and which is available under health.

A

/Live and other one is a readiness.

B

Api so which is supported through at readiness annotation. So this is available under the slash health. Slash ready, so I will just access access. This I have written something already so live. If you keep this, you will get the liveness and.

A

Ready and ready.

A

B

Let me let me show you how we can achieve that so for that.

B

So we need to provide an implementation for health check, so I have provided help health check at the liveness and other one is the readiness, so these two things what it do. What it does is, so the readiness is annotated with readiness, annotation and the liveness check is annotated with liveness, and if you you can name it like anything, and you can also send the status based on some business calculation and it's up or down based on your needs.

B

So currently I am passing with this up, but you can also do some calculation before you make the decision, so this readiness will always still have some calculation, so I'm just calculating this start up. So if you see here, it is just observing the annotation if it's a listener. So when, when the application scope, it's a it's a annotation available and when the server boots the CDI engine will initialize all these things once it is successfully initialized.

B

This observes will get some notification so based on that I am just calculating the ready time and how much time it took to get started and everything so to demo. This I need to stop my application.

B

So here I don't need to pass this anymore, so I have already demoed this to you. So I am NOT passing this so I'm, just starting my server again.

A

If I hit the ready before it start, I.

B

Am getting exception? Oh.

B

I saw something but I.

B

Want to show you something: okay,.

A

Let me cover that later.

B

So this is how you can provide the health check for your micro-services. If you see the code, it.

A

Is so simple and straightforward, the.

B

Next topic is matrix, so matrix is this specification offers a unified way for the micro profile service to export some malting data at the management agent, so it's mostly for the developers to have the uniform, uniform, follow some uniformity across all the micro service development. So they can basically depend on few concrete restrain points to check the matrix of the application. So by default it is classified under three different categories. One is the base and the other one is vendor specific and the final one is the application specific.

B

So if you access this, if you want to get some matrix at the base level, so you can pass this matrix so matrix, / Bay base, okay, I'm, not getting all these endpoints.

B

Yeah, so by default, this will. This is follows this pattern, but if you want to change this pattern, you can pass this.

A

B

Yeah, so it will be transferred as a JSON format. So you based on your business, need you can change the response pattern and you can also consume those API to get the in this. Some.

A

Insights about the underlying heap and all those information so.

B

I'll just call the application level matrix okay, so so these are the matrix. I am getting so how many times the particularly APA is called, and all these things is possible through this matrix. So you can also get some intelligence and you can also do all the metering stuff based on your and need.

B

So these things is achieved through a simple annotation.

A

So if you go here so.

B

These are the annotations is used for so one is the counter and the other one is metered and other one is timed. So, based on your requirement, you can define what type of metering you need and you can monitor your application. So this is something for the unit for calculating the metric analytics.

A

Okay, so I hope.

B

This also be straightforward to you and the next one is fault. Tolerance. If you take any micro service, the fault tolerance is one of the fundamental expectations from the industry, so meaning fault tolerance is nothing but fail fast. That's what people are claiming that if you and reduce.

A

B

Of your micro services so based on different strategy of your fault, tolerance policies. So by default they are supporting five different fault, ons capabilities. One is the time out and retry the other one is fouled back and the next one is circuit breaker and the next one is bulkhead. So these are the different fault. I mean the fault, tolerance policies, and you can you? Can you can club more than one or one based on your requirement?

B

You can do all this magic in your application and you can handle your application in an efficient way, we'll see one by one, how we can achieve this in the real time. So let me start with the timeout. Timeout is simple and straightforward. So in some situations, if application is waiting for so long and this may impact the performance or the other requests in the server, so in order to handle this effectively, you can also enforce the time owed at the class level or timeout for each and every resource level.

B

So this is very simple and this will have few properties so.

A

I'll, close and I will open a.

B

Few tolerance API, so this is a.

A

Simple tolerance, API I have created for each.

B

And everything: okay, the first one is timeout, so the timeout I have by default is thousand millisecond yeah. So that's how this configuration says. So I am not passing anything so the default will take in the place, so I am purposefully waiting I mean just to stipulate that I am purposefully, make my application to wait and then how the timeout is taking place and it is responding.

A

A

A

B

Okay, so this is responding properly because this is falling under within the scope. So this is responding based on the success. Successful response to defend here so I am passing. I mean I'm, increasing the threshold I'm just forcefully making my application to wait.

B

So this this this will throw a timeout exception, so I'm still I mean seeing this in the console, but let me show you how you can handle this exception by accommodating the other annotations in the future, but right now this is just to explain how and what will happen when the timeout exception is arrived. So when there is an exception, the timeout falls under place. It will throw a timeout exception. Okay,.

A

B

But in real time we need to handle this exception, gracefully right, so for that we can attach some of the other capabilities offered by the fault tolerance. So we will see in this upcoming examples. So the next one is retry, so here try mode along with retry. So retry is nothing, but when the retry mechanism should take place so I'm I'm indicating that whenever there is a timeout exception, just try to read, try to execute Negritude one more time all based on this come in configuration.

B

So so they have provided all these configurations just to make sure that we can extend this all these capabilities in a nicer way. So this indicates how many times this retry mechanism will take place. So if I mentioned this one, then it will retry once but I have mentioned this too. So it will read, execute automatically two times and then, if there is no I mean successful response, then it will fail again, but in between, if I am able to recover my system, then I will get a successful response.

B

That is the whole intention behind that. So I will just see how this behaves. So this one has an endpoint with retry on.

A

Okay, so I will scroll.

B

Down a little bit.

A

Or I will stop and start heads.

A

Yeah, it started and.

A

A

Okay, there are two times sorry, okay,.

B

So I'm I'm getting the successful response here. If you see the console and it is enters to the method and it the sleep time is 1 minute, so it is giving the successful response now I mean increasing the wait period. So just just look into this console so how many times it is actually retrying based on my configuration, okay, so I'm just sending this request and definitely.

A

This will fail, yeah, it failed and I will take you to the place where I stopped last time.

A

B

This is where the first one right and then this is the first request for five minutes waiting period and it tried it failed, so it trolled the time mode exception so, based on this configuration, so I have made the configuration as to the maximum retry to right so based on this configuration, so it tried two times and even within two times it didn't get any successful result, so it failed again.

B

Okay, this is how you can do all this mix and match of time mode and the retry and all these things so delay is nothing but between the red rice. What is the wait period I should enforce to the system, so it should not be very frequently you Fira tray and then the chance of success might be very low, so I'm just giving some time slot to delay my next time slot to try. So you can also configure this based on your requirement.

B

So next one is very interesting, so we'll see how we can Club tie mode with fallback method, so fallback method is nothing but to improve your efficiency of the success rate whenever there is an exception. So, for, if you see here the fallback they they offer two different flavors one is the fallback handler and the other one is fallback method, we'll see what these mean and how we can get the benefit out of it. Let me explain: first, the fallback handler.

B

Okay, so fall. If you see here the fallback handler so I have configured some class here. So what does that mean? So we need to pass one fallback handler implementation, so this is the inter face. I have provided one implementation, so this is a fallback Handler, and this will provide some kind of response whenever there is an exception, whenever there is an exception for this method, so for time owed. So this this this timeout exception happened and then this fallback will be triggered.

B

So what this fallback will do is this will have some kind of response, so I have currently generating some dummy response, but in real time business use case based on the error are based on your requirement. You can change the response and you can send it back to your client, so we will see how we can see this in action fall back and time old fall back answer.

A

B

So I'm just changing it, so first there is no exception, so I am getting the successful response from my method itself. So if you see here so this is a response.

B

Timeout with fall back handler right, yeah timeout, with fall back handler, so I am getting the successful response. When I increase the threshold, then the response will be changed.

B

If you see here, the response form from chilly dance from bar fall back handler, so meaning whenever there is a time-out exception happen. This one gives the control to the fall back handler. So please handle this exception. How you take this decision so the the chain, the the request chain, will follow the fall back handler. The fall back handler actually generates my expected response.

B

So if you see here the same result, I am able to see here as well, so this so you can improve your application efficiency by accompanying both time mode and the fallback handler. So the next one in the same fallback bucket, is fallback method. Instead of fallback handler as a implementation, you can define a separate method and handle the responsibility to that method.

B

So if you see here so I'm, just mentioning that fallback method- and this is a method name- and this method is available here as a private matter- got it so in a so, this is again a simple and straightforward, so we will we check how this operation works.

B

And I will change this first I will get the successful response, yeah timeout with fallback method, and then I will increase.

B

Generate it from on fallback method, so this is how you can handle the responsibility to different method are different handlers during some exception, during some exception, in order to improve the efficiency of your application. Okay, let me.

B

Let me cover the circuit breaker again, so this is one of the important topic when it comes to the fault tolerance circuit break, you all know what is secured, and by default, this circuit will be in the closed state, meaning all the requests will be allowed to reach the appropriate. The request, Handler and the response will be sent back to the client.

B

So when there is a problem, what this system will do is it will open a open circuit, meaning no further requests will be allowed inside to access the resources and based on the configuration there are different configuration. They are supporting right now and based on these configurations, so you can make this application to half open the circuit and then based on few other configurations, you can again close the circuit.

B

Then all the requests will be allowed inside to access the resource so we'll see what all the different properties is being supported as part of the circuit breaker. So one is the request, volume threshold and the failure, failure, ratio and success threshold and the delay. So let me start with the request volume threshold, so this is.

A

B

Within that period, my expectation is for meaning within the for the failure ratio is enforced as 0.75. If it crosses this threshold, then automatically the circuit will be opened by default. It will be in the flow state when the this principle is violated, then automatically the circuit will be opened and then the next request, so it will, it will delay for one second and the next request. If it is success, then it will monitor internally- and this is the temporary request- successful request allowed to access the resource in the ass open state.

B

So once this threshold is also reached, then it will be completely closed and then the circuit breaker is not active. So all the requests will be allowed inside. So I.

A

B

How I can simulate I tried, but I couldn't get the response properly, but I will check again. So this is the circuit, breaker method, and if you see here, this is throwing the proper response back response from the circuit break and I'm increasing the threshold threshold.

B

So my my if you see here so I, have also attached a fallback Handler along with this circuit breaker, so I'm catching here, a circuit, breaker open exception, handler I mean meaning I am cashing the catching the circuit breaker open exception. So whenever there is a circuit, breaker or circuit breaker open exception is thrown. This will be activated.

B

So that's how this is indicating okay. So if you see here, I'm cashing this fallback handler, okay.

A

B

The next one is in this category is bulkhead. Bulkhead is nothing, but if you want to enforce the thread, pull efficiency, the country and all those things, so you can have the bulkhead annotations and you can manage these things in an efficient way. So currently, the micro profile offers two different way of a bulkhead capabilities. One is the thread pool level isolation and the other one is semaphore isolation. The first one includes a bulkhead as well as the asynchronous meaning. This will calculate both the maximum concurrent connection, as well as the waiting queue.

B

The second one will only consider only the concurrent number of connections so unfortunately, I have the implementation, but I couldn't have the test to reproduce that so maybe in the future. I will demo this to you, okay, so these this. This is where this is where you can define the bulkhead file, so meaning five concurrency is alone and Here. I am just configuring annotation, as well as the bulkhead, with waiting to ask you so here you can also, as I mentioned to you, you can also mention the waiting task queue.

B

Size is 10, so the 10 records will be in the queue and once this one is successful and then this Q will be processed so that so this will work. I hope you have some fair understanding about the fault tolerance topic. I will move on to the next one, which is the rest client.

A

B

I am seeing we have lots of client libraries for the fault.

A

Testing our risky api's right, so we have apache httpclient.

B

Even java recently introduced from java 9, the native HTTP client, and you can also write the HTTP from the native URL connections and everything, and there are plenty of HTTP clients to test our REST API. But when it comes to the real-time API testing, there are plenty of problems: marshaling, unmatch link and error handling. All these are complex problems, and if you want to validate some of the response, you need to pass the JSON and you need to do and you need to have all the curry and everything.

B

But what is the beauty of the risk client in the micro profile? It is a type space type safe rest length, meaning we we we complete a Java type system, so you write application completely based on the type semantics. So you we get the opportunity to test our application based on that types. So it is so easy and straightforward to test our application as well, because that is the one of the best approach.

B

When we develop some application in the Java, because then our code should be testable and we should be able to efficiently test and easily validate our application State. So that's the beauty of the and I mean micro profile eristic length. So these are the two patterns where you can configure the Wyss client into your applications. So one is, there is a wrist sign builder and through which you can create a new builder, and you can pass a base URI, which is our application host and port.

B

And then the beauty lies here, some a period class, meaning this is the actual API definition interface. So this will return the actual type and we can simply call the method and test it.

B

So that is that is that's where the beauty lies in the next way we can test our I mean the rest client is we can directly on at 8 our interface or the API definition with register with rest client with the base path, and then we can just inject at-risk land into our code and you can call anything any method available in the interface, so we'll see how we can test some of the API. Using our core I mean there is client, so I have written some sample test goals.

B

So I'll take this one, for this. I will also open the API definition.

B

Ok. So this is the API definition, and this is the API test code. So, if you see here, countries is the best I mean countries is a path, and this interface is I country API, and this offers three methods. One is all other one is add and the other one is get and if you see the client call, my base path is HTTP localhost, 8080 and I'm initializing.

B

My res client, with the rest client builder, with a new builder and I'm passing the base, URI I, see from the URL builder and then I'm just building with the interface I country API. This returns the I country, a pin object, so from that moment, I can simply call the method directly and getting the response back. So we don't need to write any marshaling code or run marshaling code. Yeah everything will be taken care by the underlying implementation engine, so that is the expected expected outcome from the micro profile specification.

B

So whoever implements the micro profile are over, claiming that I am compliant with the micro profile. So all the framework vendors has to provide this kind of marshalling and marshalling capabilities. Out-Of-The-Box, so I'll just agree give this test.

B

B

Sometime, we need to be patient when we test our applications.

B

A

B

So this is returning a proper result so and if you see here so I'm, just calling da P it directly so country AP a dot ad country and I'm just passing the beam and deciding that so I'll just call this test.

A

B

It is also added now yeah. The link is generated, I a say, told earlier, so if I call this again so the Italian should written.

A

B

Yep, so this is also available here, so this is how we can write a simple I mean we can cover all the client, the API, using the newly introduced risk client in the micro profile. So this is so simple and easy for developing our test framework and we can cover the code in an efficient manner. We can avoid all these things, whatever we have in the right now. So this is one of the important component as part of the micro file.

A

B

Next, one is open, API I believe you all know open API. So this is a specification for writing your API definition documentation or how you can define your EAP contract and how you can share this to your developer. So it's basically when you target the API first development. So this is the kind of contract where you can define how my API should look like and what is the request it should pass and what is the response? It should return back and what is the status code and what is the payload? What is the schema?

B

All these things will be covered as part of the open, API v3 specification, so how this is coming to the picture of micro profile context. So, by default, the micro profile specification defines whenever you access, slash, open, API, the micro profile implementation vendor has to provide the open, EP specification. Out-Of-The-Box such a beautiful term I mean definition is defined as part of the specification.

B

So we don't need to write a single line of code to generate open API, but if you want to improve open, API spec, we need to do some customization here and there I will explain in a moment how we can do all these things, but just to get started with the open API definition. We just write the JPA metadata using the jax-rs annotations. So from that annotations they are automatically generating the open, API spec. So that's how the default part is being covered.

B

So if you go here and I'll just change this to open the API.

B

So by default, the response is in the ml format. But if you see yeah Mel is a problem, then you can also change that by passing the accept header.

B

Yep now it is change the JSON okay.

A

B

There are two flavors, one is the amel format and the other one is the JSON format. You can get the benefit, whichever way you like, okay, so this is by default. This has to be supported by the microbe profile. Implementation vendor. So acid developer are as a.

A

B

A company adopting to the micro profile a concept: these are the default benefit. We usually get that without any effort. The next approach is static file, so I'm I have heard some of my friends is also doing these kind of approaches, so they are generating this skeleton micro profile and they are hard. I mean handwriting. All these definition are changing the definition based on their needs and they can ship it the customer, so in this column, I mean in this context. If you take this approach, you can generate the open.

B

Api specification are, if you can create, are by hand, and you need to put it under the metin's and once once it is available, then the moment you call slash open API, then this static file will be picked up and share it to you.

B

Okay and the final approach is, if you, if, if you are addicted, to write some code to generate all the micro profile, I mean the open, APA specification on demand, then you can also do that. So the specification provides some capabilities to do that. So you have to write an implementation for OAS model reader and this needs to be configured with the key MP dot open, a pH dot model dot reader in the configuration property. Once it is done, then your the dynamic code generator will come into the picture and it will do its job.

B

There are three different flavors I explained, but in real time all these things will be merged together and it has its own algorithm to identify where and all it is available and it is taking the definition and it is to you mix and match, and all these things and finally generating one nice open, API definition based on our business needs. So we will see how we can do that. So I have written a simple, open, API custom implementation, so I will show you. So this is my custom implementation.

B

So by default, I I don't have the info information, so what I did was I'm just generating the info information for my requirement. So for the time being, I will comment this.

A

B

Here I will: what I will do is I will change the micro profile concentration yeah, so I will comment this okay, so meaning so this custom implementation will not be inactive. Okay, I'll just compile.

A

Yeah I'll start.

B

Yeah I will access the open API again.

B

So the info change, if you notice that so it's it's, this default info generated by the definition, but I don't want this, so what I will do is I need my custom info, so I have added some contact info, so this thing's is not available. If you notice here, the contact information is not available, but I want to include my contact info and everything so how this guy changes. This definition.

B

So it will consider both the default open, API spec, as well as whatever we provided through the custom voyeurism generator. So these two things will be considered and it will be merged as one and it will be given back to the end users. So that's that's! A beauty of the open API specification as well. I in the micro provide I'll quickly rush, because we are closing to the end.

A

Okay, pay attention to this statement.

B

Right now, so, this info has only two title and the version the moment I test this again, and it also includes.

A

B

A

B

My code right, so this is one of the enhancement you can do if you want to beautify your your APA and everything so, and apart from that, I have also made some changes in the below method. One is the pink method right, so you can annotate.

A

Yeah, it is very tough. Oh sorry,.

A

So I'll copy the response and I will rest here.

A

B

Want to show the pink method. I have commented, something.

A

B

Just pay attention to this okay, so this is another customization I have added through some of the annotation possibilities. So if I open this.

B

Yeah, you can also enhance your APA definition by annotating these kind of enhancement into the open, API definition. So these things is also part of the micro profile specification, so operation, so I am rather high, have added then a nice description and the response code with the proper description and everything. So this is how you can enrich your APA definition and deliver to the customer.

A

B

So I have covered no open API, so the next one is open tracing. So this is again on important component in the Micro profile specification when you are in the development of micro services, so where, if you assume that if you have multiple micro services and if you want to understand how the request is going to a different micro services in a real-time, so this will be very complicated for the cloud scenarios so to address that the micro profile offers this capability and the tracing life tracing based on the requests happen.

B

So you can also monitor based on your requirement. So currently there are two tracer implementation is supported by the Collegium framework, so one is the Zipkin and llegar, and for this demo I am considering Zipkin alone, so I am not having any demo for that, but I'll use the Zipkin and how we can get the benefit from the Zipkin okay.

B

So for so, for that, do you need to download the pin from the internet and I have already downloaded, and it's a simple Java file and you need to start these chickens over I will start this.

B

So so there is one change you need to do in order to bind Ripken with your application, so Zipkin file by default will be running on line four double one and this information which application has to be monitored and everything that has to be informed in the micro profile configuration file. So if you see here so I have my tracing service as Haley Dunbar, so this Kelly Dunbar is the application name.

B

So if you see here so that is the application name see, so that is the application name, so that is configured as the service in the tracing. So from that moment, when I start the server, this connection will be made, and so now this routine is started and I will also start. My halidom server return. Micro service is getting started and the moment it start. If you notice, in the command line console there will be something yeah Zipkin tracer builder, so it is watching on this Halligan bar. So to access this Zipkin.

B

B

We need to access.

B

B

Yeah my Zipkin is running successfully and currently there is nothing and what I will do is I will hit some end points.

A

B

A

I will also call welcome.

A

B

Called to services and I'll just refresh now, my service is picked up here and I'll search so see so here I mean request is being monitored. So one is this: so you can track what is called where when is called so you can.

B

You can track complete trays of your request, so I am currently running one micro services, so this is not very I mean interesting, but when it comes to different micro services and the requests how it is being spanned across different how the tool chain I mean their request chain is going on, so you can track all those things in an efficient way. So if it is an error, I will I will purposefully fail. This.

A

Okay, tolerance.

A

B

It's failed, so this information is very interesting, so it will be shown in a red color, so go here, so you can understand in a detailed way. So all these I mean the analysis on this tool is so efficient when it comes to the real-time. If you want to track in a distributed, the request span.

B

So this is all about the open tracing. So it's simple: there is one configuration you need to bind in your micro profile configuration file, so the tracing dot service with your micro service name. So then the application will automatically the halyard on is automatically identified the Zipkin and it understand that okay, so I think I'm I have reached the end and thank you everyone.

B

This. The whole project is available in digital under.

A

This URL- and this is my blog and I.

B

Have written many articles about the kellytown and micro profile, so you can go through that so one before we move into the question and questions, if you have any questions you can ask, but one disclaimer is I, may not have all the answers right now, but I will definitely find an answer at later point in time and I will share to you.

B

So you can get the benefit of the micro profile, easy and efficient way, so happy developing micro services using micro profile standards as well as skeleton framework, so till now, I have not used any of the code specific to gelatin framework. That is the beauty, because I have coded everything with respect to the specification. So so, whenever your company decides ok, this contract is the licensing terms and condition is not working anymore. So we don't need this framework.

B

So if you want to move something to a different framework, so you can pick and choose whichever micro profile compliance framework and you can move the complete application to that. You don't need to change any single line of code. So if I want to move this code and run into some other micro profile complaint, the container I can simply run because I have no dependency with the halida. But Helly done is a framework offer some capability, so you can get the benefit.

B

So apart from building your application, so they also generates the docker file out of the box. So I told you in the early, so they have generated the dockerfile by default, so you can get the benefit and they also offer this kubernetes deployment structure the abdominal so you can get you can deploy into kubernetes. Also so.

A

There are plenty.

B

A

B

To micro profile as well as skellydun, so please I would recommend you experience the whole frameworks and the specification, and you also contribute to the community yeah. That's all I have today.

A

Thank you so much everyone for joining into our livestream today. Thank you, Tom for a guest speaking and I'm gonna end the broadcast. You guys can watch this uploaded on YouTube channel tomorrow. Thank you. Thank.

B