AZ-204 Certification Exam Preparation Workshop | Azure Developer Associate Exam | Easy Study Path

joining in thanks qasim for having me over here for this session and uh i think it's time let's get started so today i we will be talking about 204 and this is mostly a certification preparation uh session so what we will do uh we don't have enough time with us but whatever we have we'll try to put that in a very interesting manner i have a case study with me and i will be uh walking you through the case study and i will be explaining the scenario i will map it with your certification as well that why should you do the case study which i'll give you as a lab also if you want to do later on uh if you do that how much percentage of the of your certification will get cleared just by doing that and it's going to be very interesting as well so you will like it so it's called as modern application like in today's world we are not talking about those ntr applications we are talking about more cloud native micro services or maybe any data ai oriented applications where everything comes together modern app so how do you create modern app and how do you basically design that or what thoughts are there when you make that so we have a customer scenario we'll go through that scenario and then we'll try to understand the requirements plus the solution plus how does it help us out in clearing the two not four certifications okay so i have prepared a little bit of slide deck so let me please walk through that uh before that my name is shivam sharma i will just quickly yeah this is my linkedin profile you can go and find me over there i'll put this in the chat window so i can see the introductions in the chat thanks for it i work in aws i also have 104 ac 900 i am joining in for michigan working on sql database but new to azure not using it right now sorry so this is uh my name is shivam sharma basically i i'm working in microsoft azure since 2011 now 11 or 12 like that and since the old azure portal and uh i also work in gcp i have some more understanding of aws as well my most experienced lies in in azure and now i have end up working on multiple services in azure like in infrastructure in designing architecting development on azure data platform ai it's been a long journey on the cloud so uh it the cloud sort of is like that one single uh you can say shop for all your requirements so similarly i also got the opportunity of working on multiple platforms so my linkedin profile is there in the chat window in case uh you want to join me over there you can connect with me as such in case you have some questions later on feel free to ask so what we are doing today we are let me just walk you through the what is this developer certification what do you have what do you expect uh and so and so forth then we'll uh continue with the with the case study i have i'll walk you through the requirement and the solution of it and maybe uh i'll give you the the uh the workshop to do as a uh maybe as a homework where you can deploy it as well okay so now check this out now uh this certification uh uh the easy to not for soap you have to clear one examination to get one certification it's like one to one mapping over there so the certification name is called as azure developer associate the examination which you have to take is called as two not four like that right required examination sometimes in azure to get one certification you have to clear two certificate two examination like in it happens in the architect uh track like you have to clear uh three not three three not four then you become the architect on azure but in developer because this is an associate level certification so it's not that tough uh and only one examination is what we are supposed to take not two now what i personally like whenever i do any certification uh what i do i like to read from their documentation like documentation is mostly is the most updated thing which they provide and then then the second thing is hands-on if you do hands-on you will be able to clear it because uh it's more like you 100 times you read something or versus you do the hands-on once and then you are able to understand the concept way better so what personally i what what if you ask me that do this certification then what do i do i basically go through the the details as such so the outlines the details outlines are over here i'll give you these links in the chat and i'm giving the landing page in the chat recording is anyways with you so this is the landing page in the chat from where you follow it through now the details are available on this link over here and if you want to do the hands-on part what microsoft did they created ms learn so you see this this these links are there at the bottom of the page like create serverless application connect your services store data or deploy the websites or in virtual machines manage resources in azure so this is uh this redirection happened to ms learn so ms learn i'm pretty sure if you work in azure you are aware of ms learn if not if you if i just click on this browse let's say browse there are around 3000 uh tutorials are there some of the tutorials will only have theory over there will not have any hands-on but some of them will have hands-on and some of the tutorial even give you the environment to do the hands-on in for example if you don't have azure subscription then how do you uh how do you do it so for example if i just show it to you let's say one of them which i recently used so i sort of remember it check this out so if you do this cosmos db lab over here they will give you the sandbox environment see activate your sandbox so 10 sandbox per day can be activated just give me one second so i just wanted to sip some water okay so this is that sandbox environment which you can activate from here and you can do the hands-on as such so this is this ms learn is somewhere where you should be doing your you can say where you can you can have a multiple tutorials are there which are very well written which are very you can say five star type of tutorials and which you should will help you in doing the certification now you can come to this page and you can just say i am looking for the developer track so you you go here and you click on developer and then it it sort of you have these uh 999 tutorials are there for just for the developer track it covers from the fundamentals to the end of it and then we can also do a type of learning path like i am looking for uh i'm not looking for individual tutorial i am looking for a scenario as such so then see there's two hours of scenario uh another two hours of scenario two point four one uh minutes of scenario one point four four one like that one hour three hours of scenarios like that eight hours of scenario and this is this is gonna be a long one but you can definitely do that so this is where you should be uh you can say uh this is that structured content when you're doing self-learning see best learning is self-learning so you can definitely go there and you can learn on your own as well but the main thing is you have to do hands-on now when you go for the certification there is also some official uh content given by microsoft now microsoft has gone uh sort of open uh now so what they have done they have uh released everything on github so if you when you are trying so for example if you go to any institute and or as such or you come to me for the training as such even what i do i i basically walk in the training we do the official labs so official labs are listed on this github page i'm giving you this github page in the chat window now if you are doing self learning you basically go to this page there are 12 labs over there tutorials are there these tutorials are not easy some of them might be easy some of them are pretty difficult so you'll you'll use to get used to like polygon data and all that it's pretty long so what you can do is you these are those uh 12 uh labs given by microsoft officially the previous ms learn was not directed only to the certification it was like if anyone wants to be a developer in azure then what all should they know but these 12 labs are just directed to the 204 certification so you can clear the developer track as such right so uh i recommend that those who are planning on getting certified this is what you have to hit first you click on them and you deploy them in the portal now you need subscription to do it if you have the subscription amazing if you don't have the subscription you go for free trial right before anyone ask you go for this free trial sub subscription just look up uh just google for it that's the free trial if you have already used your credit card get some credit card from your family members this this subscription is free of cost but you have to put some credit card in the starting to be able to uh get the certification get the subscription in the starting itself so uh because of that if you have already used your credit card i used mine i think in 2014 so uh then i sort of use other people's if i need free accounts i use other people's credit cards over there okay so and just the good thing is whenever the word comes credit card it's very risky right what happens if we forget to delete the resources and it starts in the credit card nothing like that will happen when you're when the subscription will expire they will give us 200 as such when the 200 will expire the account will automatically block it will not they will not charge you a dime from your credit card it's just for authentication purpose whether you are you or you are some sort of bot attacking you can say attacking the uh azure platform so uh that's where you have to uh from where you get the subscription these are the labs which i have given over here which you will you will follow after this session to be able to uh get certified right these two things are there in the chat window uh keep a note of them now what we will do i'll walk you through the outline as such and then i will walk you through the case study which i have and then scenario we will design this scenario today right and with that scenario we will pick up uh the topics from this uh okay you can say from the tutorial from the uh outline of microsoft's 204 certification and we will try to learn something in the session that's what we are planning uh just make sure your cameras don't get uh enabled without uh uh yeah you know it got enabled okay so uh web app is something which is one of the uh major section before i go you go in this see they have given another page here so in this page what happens they give the percentage like 25 to 30 percent will be compute solutions compute azure compute solutions right when you talk about azure compute solutions if you uh i'll just give you one very important page which i generally use personally also when i have to architect solutions not this one so in the architecture center architect uh yeah i think this is the one so see uh this page is going in the chat this is basically the design tree or the you can say the the dag diagram over here now if you know these services you are good right what is azure virtual machine batch app services function container instances service fabric aks uh you're good if you know this then the first part is good like this part is pretty well covered you know more than what they will ask in that particular 25 to 30 percent of the certification if you don't know this this is you can say you have to cover that part there is a right and then you basically will be good enough to get hit that 25 percent when to use what uh if you upgrade that that's on this page like which technology should you choose in which scenario why should i do this that y is on this page it's in the chat window so 25 is basically compute solutions uh in in that you have to talk about acr you have to talk about aci uh we have to talk about images erm templates and so on so forth web app azure functions are there uh 25 uh 15 to 20 percent will be storage solutions right so this is where cosmos database comes into picture cosmos db is pretty interesting that's the one of the most interesting solution in the clouds in any of the clouds today uh the the global distributed database which is providing the five type of consistencies plus it provides concurrency plus it it provides multi-modal uh design like this is a semi-structured database one database which is equal to cassandra one database which is equal to mongodb one database which is equal to uh sql uh you can say not really relational this is non-relational but it is it provides the same apis so cosmos db is something if you are new to azure i recommend hit cosmos db first if this is one of the interesting solutions you will like this more than the other solutions uh so uh start start your things from something which is very interesting right so start from this and learn learn it up like what is cosmos db details of it uh why do we have advice so important what is the homing apis over there and so and so forth a little bit of cosmos db we will discuss today in the case study uh blobs are the old blobs go through it uh the object storage uh go through it the uh then security is something which is very important in today's world without so it's when we design solution it's security first it's security is not implemented afterwards it's from the day number one we implement a security everything has to be secure multiple layers of security multiple solutions in azure network parameter application data uh the when it's it's more like an onion right so security is more like an engine if the more the number of layers you have more your security environment is so it's harder there's a call there's a concept called a cyber kill chain so it's uh the cyber kill chain becomes bigger and the the ability of these solutions in azure to be able to detect when somebody attacks also becomes very you can say more probable when you have a multiple layers of security so azure ad is one of the core identity management service like iam where all the username and passwords are kept but that's the easy way to describe azure 80 usernames passwords but it does many more things so uh this is uh azure 80 is definitely has to be used over there what's it what is shared access signature uh identity platform securing the cloud what is keyword this word you will see it doesn't matter which company you work in uh if you if you're properly architecting your solutions keyword will be there without keyboard it doesn't work out right your application want to talk to database your uh big data cluster wants to talk to the database your uh the api gateway needs one ssl certificate to be binded to it everything will be there inside the key vault and then it then then we bind all of that so key vault is is not very important from the uh you can say from the certifications point of view but key vault is very very is uh important from your practical productions point of view it will be there right so okay troubleshooting and optimization optimization is uh we all want to save money right we all don't want that things are running we have overprovision resources and resources are basically uh you are paying for something which you are not using so try to use serverless try to uh reduce the resources scale up only and if you provision the resources you make sure you are using the resources right secondly how can you make things faster uh if you have so there is a concept called as scalable web application so i'll give you the architecture of it this is something if you uh web application like that in azure platform so over here so this page is pretty important if you understand the single architecture and all the components which are there it will uh you can say a very big big percentage of the certification also gets covered and a big percentage of the of uh practical modern scalable responsive applications in azure also gets covered with that so if you see they have azure dns over here for the name resolution they have active directory for the authentication purpose as your front door is called as application delivery network it makes it accelerates your net your application the the traffic reaches the application way faster cdn is the static content delivery like the reverse of it when you're pushing the videos and and images and all that back blob is the cheaper blob storage uh web app is a web pass solution cues will decouple the web app layer this worker role with the with the this web role with the worker role over there this queue will decouple so how many types of queues are there in azure so what you have to go through is that uh how many types of uh uh queues are there in azure when should we go for which one for we have service bus queues we have storage account queues we have service bus topics as such so in which scenario what has to be uh gone through so what is azure function how do we use it is it so this is serverless so is serverless the uh in which scenario we do not use serverless servers everyone knows it's pretty cool and all that but when when should we not use it right so uh that you can find on this page uh when should we not use it and secondly when you use serverless the problem is serverless components they scale very fast they scale very quickly they scale to number of instances so when that's high scalability happens the practical problem is that the other backend services are unable to handle that scalability if if this layer scales up my databases will not be able to handle that much of request which is coming so i cannot let it be scaling to uh infinite as such what i can do there is no infinite but practically you can say maybe 100 200 000 1500 instances but that's a huge number right so what we do we in serverless we always put a upper limit so in this azure function we have the host.host.json files are there in those host files we put the upper limit like that when you are scaling don't scale beyond this limit because if you do that my backend systems will not be able to handle that and i will end up having more trouble than what i was trying to solve then what is this cache doing over here there's a concept called as cashier side pattern you can watch any of the uh previous uh i i remember there was one one session on the design patterns on azure i think kassam himself did that so you can check out in that session what is cashier site pattern and how do you implement that with the with with the databases what is at the read time what happens at the right time what happens so how is this uh making sure that the uh the responses are coming way faster uh azure redis cache and secondly when we should not use it uh if you don't have frequently running queries you don't use the cache because it will add that overhead latency and it will make things way slower than faster versus faster so uh right now uh relational databases uh this uh this is a relational database that is semi-structured right this is where you keep your structured data this is semi-structured in nature cosmos db semi structure no sql semi structure no normalization nothing like that so and global database so when should we have this one when should we have that one what are the requirements uh when we should go for them and what will this give me what will that give me right uh in in that sense so and how do you as your search is just like the cherry on the on you can see on the cake here but this search in this diagram is just basically uh doing that scrolling over the it's crawling over the data sources creating indexes over there over which the functions can can further iterate so this one architecture is uh is very useful the one thing which is missing from this architecture is the high availability like the these are single point of failures right this is a single point of failure that single point of failure so this is not high available for a high availability you can look up for any other architecture over here where they have some sort of active active uh uh you can say active active scenarios running or active passive scenario running so you can further think it through but that goes more in the architecture session it it's not uh making the high availability over there is not really responsible for developers uh making sure that's something the architecture you can say the architect should do that so but this is this is something if you go through this details of it are on this page and you learn why should we do it and implement uh if you want to deploy this all of this is available in the labs which i gave the 12 labs from microsoft they have this in bits and pieces like they implement cdn somewhere they implement queue somewhere functions are there uh cache is there sql is there and and so and so forth so uh all right so this this is where that cache is there uh caching patterns which are used should i do expiration eviction or encryption and so on so forth so this is around 15 to 20 of your uh certification uh monitoring and logging the application insights is is there since since ever i don't remember when it got started but i i am seeing this in so many years now it's there it was there in azure since the very starting itself this is one of the oldest service uh the new services azure monitor when they they combined everything together but application insights is something which you have to use uh for the monitoring of application like uh when you open the page what was the turnaround time what error came which error came four not four how many times 500 came how many times uh uh maybe 200 happened uh from where the person opened if then the application was talking to the queue what was the latency between that like was it uh is it available in uh in let's say the average latency was nine millisecond is it nine millisecond or is it more or is it less so all these details will be there uh is what application insights can get me uh i can get me from the from the uh application so this is the one of the major monitoring service which developers will be looking at the dashboard of application inside is something which you keep open and then you just keep on looking at that that it's when when if something fails this is the one which will tell us and we will quickly go and fix that so this is that service now uh 15 to 20 will be third-party uh services uh like how do you connect and consume azure like you have to apis come into picture so apim is like fpg apigee type gateway apim is a very very interesting solution this is one of the good solutions where uh which doesn't have many downfalls like sometime what happens the solutions may be having some okay it doesn't support this so azure sql pass doesn't support that or there are some limitations yesterday one of my uh trainings uh one uh even there was one another limitation of azure sql pass pass which which even i uh i didn't notice which was brought to me by the participant that it doesn't do that i was like yeah it doesn't do that so but apim is one of the services which is like other people try to copy this like let's make the gateways like apim so go through it uh event hub event grid uh event there is there are two things like event grid is used for reactive programming reaction happens like when something happens then based on that you should trigger another workflow another process or then should some something should carry on so that reactive programming is even event grid the events are are discrete they are not continuous when you ingest data on a very this is this event hub is kafka cluster it's basically at the back end is having kafka there are hundred broker thousand broker 10 000 broker uh skus are available so but basically that's that's uh that's apache kafka at the back end so it's an injection layer solution it basically ingests data it doesn't uh process or anything right and then as we were discussing what is service bus queue storage account q and so and so forth so this page is where you have the uh outline with you in percentage that's something which you have to keep in mind now what i i will do i have this uh case study with me see so i don't want the session to be without discussions as such so i have this case study with me and this this solution of that case study uh which covers uh around 90 of the 204 and it goes beyond that so uh it's good enough right so what we will do uh we'll go through the case study uh i'll explain the scenario which which is happening and then i'll walk you through this solution like uh the various bits and pieces over there how is this cosmos db doing what it is doing what is this azure function doing what is stream analytics doing uh what is this postgres doing can we replace this with this sql what is aks doing over there what is this iot hub event hub doing and and so and so forth right so that is something which we will be discussing uh today the solution of the scenario so we will understand the scenario first requirement of this customer what do they want from us and then we will look up the solution then i will give you the solution as a hands-on lab whenever you get time do it if you do it uh it's it's something which is very uh you can say you can learn a lot from that one single scenario right now before i start with this case study if anyone has any question feel free to put that in the chat window please use the chat as the medium to communicate we will take the questions in the end i'll just open my whiteboard in the meantime any question which comes to mind i kept some pictures opened okay so what we are doing is uh today we are talking about modern app let which i want as well and uh let me just yeah okay this is the this wheel will go to that that design in few minutes check this first right so uh what we do is uh we so this customer is basically called as world wide world importer so there is a customer called as wide world importer and this customer basically have seen uh it's called as cnc machines uh computer controlled uh you can say machines as such uh the the exact term is let me just quickly is is computer numerical control machines right so this customer which we are talking about is having these cnc machines so the machines are computer controlled that's the soft meaning of that you guys may be knowing more about cnc than me i just uh sort of am able to relate with this that some of these machines are there computer control machines are there and this customer is looking for one uh scenario uh as such so the name of the company is wide world importers and they have around 50 let me just read how many they have five factories with them i'm reading the scenario from my secondary screen five factories are there from the five factories the customer is getting is having uh 900 different type of you can say machines are there 10 000 sensors are there and overall they have around 50 000 sensors in all the five factories in all the machines which are sending some data to the cloud so there are around five factories and around 50 000 overall sensors which are sending the data right so here so all these machines are sending us the data see whenever in today's world data is pretty strong right data is something which is very useful so if your application is able to consume is ai oriented or is it is if it is data oriented then you may be having that edge in the market uh versus your competitors so now uh the preferred solution which we have for this company is what i was showing you earlier but let's let's check out the requirement first so 50 000 sensors are there and the data is coming as a stream uh right now right 50 000 sensors as such and streams are coming in so the application is supposed to do multiple things but the first problem this company has is that this stream has to be ingested see there is a stream uh coming in the stream has to be ingested so in azure whenever you talk about stream injection solutions we have two two things as such we have iot hum so there is one stream injection solution called as iot hub so iot hub like that there is another one called as event hub like this right the rule of thumb is that whenever you have whenever the streams are coming in from the iot devices you go for iot hub as the name suggests and whenever you have any application sending the data like uh imagine zometo apps were there or some sort of uh uber uber apps were there and some events were coming from the uber apps people were trying to book the ride they are saying my driver is not coming on time i'm getting late whatever so then we have we go for event hub right these these are two pass solutions which are there in azure platform uh which ingest data what are these these are platform as service solution pass solutions and they ingest data ingest data now when the data is ingested then only can you do something with it if you are unable to ingest the data uh properly then your solutions will not will not behave uh like that so whenever you ingest data you then then comes the processing part like how do you process it so in the modern applications they might be using some injection layer solution as iot hub or event hub to get the data in then we start doing the processing over there see now we have a velocity problem the data is coming from 50 000 machines at the same time right in some streams what type of data might be there uh temperature data vibration maybe noise as such maybe any uh human input happened uh maybe there was a pressure uh there was a some sort of uh gauge was there somewhere some pressure was there which was used to move the pump around the pressure got released or anything as such is is what we are thinking as the data which is coming in so in the modern applications which are somewhat data oriented what this wide world importer want us to do is they want to have a very scalable consumer so see you might have heard about publisher subscriber pop sub right very interesting words pops up like that so pop in the sub so the publishers are these devices right now these subscribers are there in azure right and the broker is here this is the broker so these uh and we just talked about azure function now what azure function does azure function if i just come back to your certification azure function is serverless in nature right server is different is different from pass in serverless what happens we have compute on demand if you don't use it you don't have to pay for it so if you so what i'm showing in the screen right now is that azure function is scaling out as the data if if there is more data more azure functions are there if data is you can say stream is there or not there then they will scale in so they will scale out and scale in based on the load as such if you are not consuming them you don't have to pay for it this is not the case in in platform and service solution in past solutions even if you don't consume it let's say you allocated 14 gb ram but in that point in time you are using 11 gb ram then you are still paying for 14 gb ram you are not paying for 11 gb ram that doesn't happen in the azure function we have a consumption based plan the downfall of consumption based plan is there is a concept called as cold start cold start may happen so to avoid cold starts some of the companies what do they do they don't use the consumption based plan which is supposed to be used and they don't go for the serverless they they in azure we have a concept called as app service plan they go for app service plan and they put the function inside app service plan which doesn't make sense to me but they they do it and it's no longer serverless but yeah they they don't have the cold start problem so cold start problem it means that if something scales down to zero instance and it starts next time it will take two three more seconds to start for the first time that's called as cold start so if your application is so sensitive then cold start you have to avoid cold start otherwise cold start won't be a problem for most of the event driven applications as such so whatever data is coming in this function is the is the subscribers or the consumers which are scaling out and they are consuming the data right now right now uh as such now what we are doing behind this uh over here for so i am designing the solution on the way while explaining the requirement to the to you now if you you were talking about the cosmos db remember we're just discussing about that so now cosmos databases is a multimodal global database meaning it will have multiple endpoints you can select the endpoint uh so imagine this dotted line is the namespace and inside that you have three uh regions in azure where you made the cosmos db so over here so this this cosmos db is in east u.s this one is in southeast asia this cosmos database is somewhere in south india let's say over there like this and they all are syncing with each other see they all are syncing with each other in real time real time means practically whatever data size you have and all that is based on that cosmos db is syncing so what is cosmos database cosmos database is first of all semi-structured it uh there is no table there table concept is not there we have key value pair that the cosmos db tables are key value pairs so semi-structured nosql uh distributed database where multiple endpoints can be there there is a concept called as request unit which you have to understand uh in when you certainly get certain for your certification so request unit is how you pay for the cosmos db for example in cosmos db uh like if you write 1 kb per second right if you write sorry if you read 1 kb per second this is equal to 1 ru so 1 ru is the amount of resources required to read 1 kb per second but what happens if i write 1 kb per second right i don't read 1 kb per second i write 1 kb per second then the approximate ratio is 5 values so it takes the rights are 5 times more expensive than reads which makes sense because rights are expensive than reads anyways so cosmos database is good for many scenarios where your schema is loose so when should we go for this global distributed database when when my customers are globally distributed secondly i don't know the schema schema may change on the way today you have four columns today tomorrow you have one more entry coming in null values might be there loose schema is what you're looking for is these are some of the examples of when you go for semi-structured database so when you will deploy the scenario i will give you you will also use cosmos db as the database where the functions will be interacting with so functions will talk talk to them over here right so functions are interacting with this cosmos db like this like like this over here so if i uh if i just take you you can say to the the architecture uh back so i don't have to manually make it but now you will be able to follow it with me so this wide world importer was getting those uh iot sensor data is there and basically from 50 000 devices the data was coming in and we have the iot hub in the starting and where the azure function is the consumer behind that and cosmos ddb was the database which was being used at the back end now i will just go once one step extra uh when you work with this cosmos db right in cosmos db we have two things inside that we have the oltp part as well and we have the olap as well so olap is are the systems which are running the business oltp is the sorry oltp is the systems which are running the business olap are the systems which are analyzing the business so cosmos db is so interesting because it runs your business at the same time you can do analysis on the business so this this is something which all the companies uh you can say are looking for and the short form is called as htap hybrid transactional analytical processing systems like that so cosmos database has multiple use cases so what this wild word importer is doing they want that like in the end of the day they are looking for that mobile application or they are looking for that web application which is where if something goes wrong that what goes wrong should be uh should be should be shown uh or are the floor managers those who are in the factory right now near the uh the cnc machines they should be able to have a mobile app or a web app and they should get notification that this failed the the device is going to fail go and repair it or stop it or do something with it before it totally crashed down so they they want that web app to be there so this is the azure uh web app but this wide world importer company do not want to go for the native uh you can say uh any any other old school architectures they are looking for they literally have a requirement of of micro services they they are looking for that micro service architecture as such so this uh web app at the back end uh this interface should be should be interacting with multiple micro services as such so how do you run web apps so this is one of your certification questions over here uh there are multiple pricing plans like if this is azure website so let's say this is azure website now it runs in our wrapper which is called as the app service plan so app service plan now this wrapper which you have is having multiple sku remember remember one thing uh from this uh session that the hundred percent will ask you one question that if you are running one web application in production what is the minimum app service plan which you should have answer is standard you should have a standard plan to be able to have something running in a as a past solution in a web app in azure uh in production standard plan apart from standard plan in production we also have uh isolated isolated is when the v-net integration happens we also have plans which are uh which is called as premium plan so this this will have uh like there is a concept called a standard plan over there standard then we have the uh the the premium uh and the isolated like this so like this and there are some non-production plan below it but the minimum production plan is standard i don't know why but they ask this question they are asking this question since since very starting i remember this certification was called as five three two uh many many years back it was called as five three two so even that day i answered that question then i got re-certified in 532 then the same question came again then 204 then the same question came again uh it's something which they like so standard is a minimum plan which you should go for so uh okay now if you want to run your microservices in the cloud platform uh there are multiple methods by which we can do it one of the you can say go to method will be using aks so azure kubernetes services let me see if i can pull one picture of it okay so okay so this is not kubernetes services over here now uh okay so uh this is slightly where we are going little little beyond your certification uh aks so this is the icon of aks azure kubernetes services now aks is the is the go to location where you will be running your micro services you will containerize your micro services and you will run them inside the kubernetes services in azure platform now when you deploy something in aks there is a acr there is a repository where you have to put the image so imagine that this is a developer so developer wants to develop the micro services which uh which are running in azure kubernetes services so they uh the images containing the the micro services needs to be stored somewhere we need some repository for that so in azure we have a concept called as azure container registry so let's say azure container registry acr uh apart from acr you can use docker hub and all that but use acr because you're anyways doing it in azure so what what are we thinking of we are doing docker push over there so let's say this is docker push as such and things are going inside acr like that so this is your micro service and some versions are there images are there which is right now stored in this acr then what do we do from this acr we do the deployment in the a case so it deploys in kubernetes like that so micro services the images will be stored and what is this over here there's a docker image right so that two boxes which we have are docker images as such what is there inside the docker image the the literally the application code dependency and all that is there inside it so acr is part of your certification you have to understand how to interact with acr you have to understand how to uh push into acr how to use acr to be able to uh pull the image from it and then do the deployments now in your certification you uh you don't have much there is not much on aks you should just know it it exists in the world it's very useful if you know it exists in the world and you deploy it and you are able to work with it then that is something which will which will get you the job when you get certified you go and you want to get the job they will ask questions on a case no one is going to ask you questions on on anything else they will say what is kubernetes now you are in developer in azure uh you're a developer in azure uh what is kubernetes do you know how to work with it so if you know how to work with it then only your 204 certification will end up giving you some roi of the time you are investing otherwise no not really so learn learn this go go go ahead and you have to understand how to work with it second thing is we have another smaller brother of of kubernetes called as azure container instances so if when the customers do not have the micro service requirement they are saying we are not looking for the orchestration we are not looking for that uh massive amount of uh you can say networking and and everything to be done we go for aci so aci is the uh you can say literally like single instance so i if let's say if i have one micro service uh what i can do i i literally one micro service doesn't make sense but let's say you want to say i want to run one container then what should i do you want to run your one container uh deploy this aci instance give it some gb ram let's say it is having 1gb ram and maybe it is having two cpu core right and this is this is your application code here dependencies inside it and uh then which is running inside the aci so aci is when we don't have that microservice requirement when we don't have that kubernetes requirement when we don't need the orchestration then we go for this so for your certification these are very important for your for your interviews kubernetes is one of the most important thing which you should be able to work with in azure platform aks people will be asking questions on this uh world is going crazy on kubernetes they a lot of them don't even have the requirement but they still want aks to be there in the environment so they are forcefully putting it there so anyways you have to work with it so learn it now uh what we are doing uh this wide world importers is also looking for a little bit of the modern part like when you talk about modern application the meaning is that your application should be able to interact with the ai services in azure should be able to interact with the data platform services in azure and uh should be able to combine everything together under one umbrella as such so that modern part is that you can say is the data oriented part now whenever you have the data oriented part analysis comes into picture you have to analyze your data in real time so if uh so let's see over here imagine that you can say imagine that this uh that the uh the data coming from this cnc machine over there uh the there is a temperature sensor over there right and the temperature sensor gets triggered three times in that one minute three times uh or four times or five times that it is beyond critical limit so what we want is that if if any if the temperature is going beyond the critical limit we should be able to find that in real time and should be able to you can say work through it so uh we should be able to stop the machine we should be able to notify the people on the floor over there saying that this machine is going to burst uh please go and turn it off right so what do we want we want analysis to be done in real time so at times what you have to deploy is what you you will end up working with will be some sort of uh you can say streaming solutions as such you have to write some piece of code for to be able to uh do stream analysis so whatever data you are getting from the ingestion layer i'll just make a box there so it doesn't matter which you are using maybe you're using iot hub maybe you're using event hub based on your requirement let's say these two but uh at the end of the day we are basically you can say doing some analysis on it so the stream which is coming over there we have this service in in you can say in azure called stream analytics so here see this is called as azure stream analytics now stream analytics is a very interesting type of solution where you can run javascript functions you can run sql functions over there in your modern application to be able to uh you can say work with it so this is where we have those tsql queries so tsql and something like this so whenever you are working with the streaming data there is a concept which which is called as like windowing functions like whenever you develop anything for any any solution like stream analytics you have to work with windowing functions windowing functions means like the data was coming in real time and then how do i run my queries on top of it how do i run my javascript functions on top of it so let's say over here and then outputs is going on so this is where we are doing the analysis let's say the analysis over there so uh the when you develop anything which which runs on stream of data let's say one day your company asks you that okay you have to develop one uh function uh with with some uh maybe in a python or maybe with tsql maybe with javascript maybe with anything else but it should work on the stream of data we want we are looking for doing the analysis in azure platform then how do you do it uh i'll just give you one example of it so we have examples such as tumbling window see uh tumbling tumbling window like that so these are standard uh concepts in the world over here so what tumbling window does for example i have let's say uh different uh there's a timeline which i'm trying to create uh let's say this is timeline is second based so your data was coming from the iot sensors right now or they were coming coming from the customers they were the data might be coming from tweets also so this is the data points here right like this so imagine you have a window function so this window size is 10 seconds right 10 second window size so data is continuously coming you have a window size that windows whatever fits in the window is on on top of that you run the query so the first query executes on on this the second query executes maybe when the data goes and this is let's say data came between 10 to 20 second and this is my 10 second window over here right like this so then the query executes so the practical problem you will face when you were developing something for stream solution is that data is continuously coming so you have to hold data in one location for some time then you run the query on top of it then you get the output so we have options such as stumbling window where overlap doesn't happen the window keep on moving up the data keep on coming up and we keep on running the queries or and we enrich these queries with javascript functions second thing is like we you may have some scenarios where you want overlap to happen so overlap means like you have a hopping window so imagine something like this so uh over here let's say 0 maybe 10 uh 20 30 you're saying that we are not looking for uh non-overlapping data it's possible that some of the data is overlapping as well so we have hopping windows see windows over here so in hopping windows what happens like imagine there's a 10 second window right like this so now uh you are saying that hop size is five second window size is 10 second window size is going to be 10 seconds right so it's gonna going to overlap so in hopping window c what happened the hop size was five second so the window the data which was in the previous window is also there in the next window right it is overlapping so this is when you are saying find with the counts of the tweet every uh you can say of last 10 second every five second so hops are going on of five second if the hop size become equal to window size tumbling hopping window becomes tumbling window so whenever in modern applications you have to work with stream data then these type of questions will be there apart from these two we have many more type of session windows as well which may be in another session one day we'll cover so data comes in you write your functions in some sort of tumbling window manner or you write your functions in some sort of hopping window manner and then that is being doing the analysis for you so whatever analysis you do you store it so here you might be having some relational databases so in your certification you have to learn about azure sql uh it's part of your certificate cosmos db is also there the azure sql is basically available in two offerings there is a pass azure sql there is a serverless sql so yes databases are also available as serverless today uh what happens in serverless database is that the compute on demand but the storage is always there so i'm pretty sure you have an understanding of what is uh uh bass but serverless database is really new right so what happens in this server list is something like this see so there is a upper limit and then there is a bottom limit right over here so upper limit imagine 120 gb is your upper limit bottom limit is 1.25 gb like that so and this is where we are saying performance performance and performance is having that relationship with the cost if there is more gb ram allocated you have to pay pay more if that more less db ram then you have to pay less so in serverless databases the the compute is on demand like if you need more ram you pay more in that point in time but if you don't require it it goes down right and there is one mechanism to auto pause as well so if if things are not running imagine there is nobody trying to book an order in the nobody's trying to book and uh you can say uh uh nobody's trying to book the ola cab or uber cab or any cab in the morning of any uh taxi aggregator as such so there is a pause time say so the there was nothing running for this one r um there was no query running for this one r so the database will auto pause it will become zero like you are not paying for the database anymore it became zero like that right compute is zero over there and then in the morning imagine at 4 am somebody again booked the cab so this this is when the database will again get triggered automatically so this uh this way in which the data base is uh is costing us only for what we are consuming the ping part and when we are not consuming we are not paying for it is called as serverless right serverless databases on on demand compute this is something which you have to work with uh when you develop solutions with azure pass and serverless as such needs to be worked with technically platform as a service is equal to serverless in azure sql scenario but there are some limitations in the maximum number of ram which you can have in that point in time now okay so check this out so uh right so azure sql may be one of the output location this is relational in nature so you have tables over there the output locations might be something like a cosmos database as well which are which if you have the requirement of going global then cosmos database will be there now in cosmos database as a developer you have to understand the the consistencies which are there there are five consistencies i'll just show you one picture of it and the one which we use is called a session consistency mostly we end up using session consistency not always maybe but most of the time uh like this so i'll give i'll give this link in the chat so focus on these uh these five understand what do you do with them so this is when the strong consistency is what relational databases provides right they like if there are multiple databases they will they will sync first then they uh then the then the new commits will will reflect acknowledgement will come only after the sync as has happened to multiple endpoints so bounded stainless is the one where we have the uh stainless is there with along with you can say strong consistency session consistency is the one which you have which we generally use in cosmos db or which from the developer's point of view we end up using that like that's what we end up doing now little some soft point over here the these two consistencies are slow they are they are pretty slow the from session onwards cosmos database becomes two times faster than the previous consistency levels like minimum two times faster performance will be there than the previous previous ones our session is the one where your data will be consistent in your own db session uh go through this statement is very uh you can say in depth you can spend some time later on thinking it through like what is that my session uh read your own rights what's the meaning of that uh that's what happens in session prefix is something which we talk about uh you can say is is something where the order remains same this is more like what twitter uses at the back end like order remains same eventually is like eventual nothing matters uh this we generally don't use in production it becomes very uh it's pretty tricky uh this one eventual nothing there is no order is no guarantee so don't use this generally we use this uh that and this in production consistent prefix also makes sense in scenarios this one no not really if you have any low impact uh uh really not important uh application where there is no order required you can go for it but otherwise no real not really so this page is very important for you so everyone who is planning of getting the certification needs to go through and understand what's going on in each and every line over here uh look up in their documentation look up for uh what what things are there what details are there and have a uh spend some time in cosmos deep it's very useful for every company so some important topics are aks uh and cosmos db for your interviews uh right there's a question in the chat saying uh the when going serverless is there any impact of cold start so okay yeah so what will happen in this serverless sql so see at 4 00 am when this will start again right here imagine at 4 am somebody plays the order uh in the zumato app or in the food in whichever part of the world where you are joining in there must be some food delivery applications uber eats type of thing so at 4 am uh here when the query came there was a start the database started this is when cold start will happen cold start if i tell you the exact number uh the database will start in uh three to five seconds like that so database will take three to five seconds to start this is cold start in serverless literally it will take three to five seconds extra so uh uh the query will take uh more time to run uh first only once then it normally starts so this is when cold start will happen at that point like that okay so let's go back to that cosmos db okay so this is something where uh you can say spend some time in cosmos db consistency levels and work with it one more important thing which is there is uh i'll just see if i can find uh the example okay so please notice this notice this diagram which i'm showing here there are three services which are pretty confusing as well but they all are important for the certification uh this link is going in the chat as well now what is event hub event hub is injection layer solution at the back end we have apache kafka so what event hub does i'll just so uh just just give one more minute into it so you see this event hub which is ingesting the data at the back end what this does is it is having kafka cluster there literally so this this actually is having apache kafka at the back end so it looks like this so there are virtual machines running you don't have to pay for it they manage it but see we have this apache kafka so right over here so this is apache kafka machines at the back end uh there is a concept in apache kaffir called as topic so this is a topic over here this is a topic topic is that vertically partitioned queue that's that's how you can reduce the importance of topic by saying vertically partition q but technically that's what it does so it increase the capacity or throughput there so what is that event hub event how is this purple part so you see this purple part which we are creating like a box on top of it so the purple part is event hub uh and at the back end we have apache kafka so whenever the publisher they write the data the data goes in the topics as such and then these consumers they they subscribe to the topic and they read from there so here some partitions will be there see this is one partition another partition another partition let's say another partition another partition so data goes in this manner when the publisher will write in round robin manner it goes like this this is data going inside like that so event hub is is basically an ingestion layer solution which ingests data which is having this apache kafka which is at the back end and big amount of stream is coming in and so and so forth and event hub is this purple part which is built on top of the apache kafka which we don't have to work with which is good otherwise it takes it's pretty raw to work with it uh over here now but what are the other solutions if we go back to the page i was showing you earlier there were two other things which were written over there so there is one thing called as event grid and this is service bus right so event grid is used in reactive programming so what is the meaning of reactive programming let me see if there is any picture given by them okay great see so data is coming in from event hub what is this event grid doing uh i'll show you uh i'll explain one scenario little scenario here let me see where i can fit that okay i'll just fit it here so uh think it like this so let's say there was one uh azure web app was there this is azure web app and basically an image gallery application so this is the customer is is uploading their uh pictures over there so the customer uploads the picture as such in using this image gallery application so we the application part is there in azure web app now the back end right so let's say there is an upload upload button here so you can upload your pictures and then thumbnails will get generated and will will be displayed in the in the application and if anyone want clicks on the thumbnail then the app the picture will load back again so when you create something like this so i'll just explain one scenario so let's say the pictures at the back end where do you store this unstructured data object storage unstructured data will go in object storage so object storages is azure storage account containers so that's where unstructured data will go this is azure storage account so you have to know how to what is the sdk of azure storage account this is very important from the certifications point of view how do you work with the blobs the console is called as blob binary large object how do you work with it so like pictures will get uploaded into this uh you can say over here in this uh image container and let's say this is image container right now we have to create the thumbnails also so let's say thumbnails are there which are supposed to be created now who can create the thumbnails in a serverless manner azure function they are the best to do it so this is azure function here see so what this azure function is doing this uh this uh this azure function is basically doing the uh you can say is doing resize of of images creating the thumbnails out of the big images over there so this is azure function so azure function picks up the data resize the picture into smaller thumbnails put the output back over there and then this is displayed back in the app over there right this is the resized pictures over there but the question is how will azure function come to know that the picture arrived like there is a picture over there how will this azure function come to know this is where event grid will come coming into picture so see this uh event grid is one service which is there in azure behind behind the scenes between multiple services like storage account to functions between logic app and functions and so on support logic app and storage accounts and everything so this yellow part is event grid so what event grid does event grid basically uh is an event passing mechanism discrete events it doesn't do streams as such this is this yellow thing is azure event grid so what we can do uh over here right so what we can do is when the blob will get created one event will get generated call is blob create event so blob create event so whenever the blob create event gets generated over event grid we can pass that so this event can be passed over the event grid this event will not have the data over there it will have the uri of the blob so events are very lightweight we don't have data there you can put some little bit of payload into it but you cannot upload uh huge massive pictures into it right no not really so what this so whenever the blob create event will do what we will say event grid whenever that somebody uploads the hd picture you tell the function that function go and read it so this event grid will pass the trigger like that it will trigger the function and that function will go and read the picture like it will come to know using the uri which picture was uploaded function will go to the storage account will read the picture will will execute will resize the picture and will upload the picture back so this this concept of reactive programming is what event grid will help us in performing like making people aware what making one service trigger through when something happens in another service so if you want to perform what i just told you in the screen i'll give you one tutorial in the in the chat in one second if you if you do that one tutorial your event grid topic will get cleared for your certification like you don't have to further think about it you will have that in your pocket so what i just told you is what i am giving it to you right so this is their microsoft documentation i'm looking for the same example where we are doing the uh event grid part so resize the uploaded images okay so uh this is that image image gallery app right so this is your part a of the tutorial it goes in the chart over here and then the event grid part is the second part over here so where you we will do the event grid and resize with azure function so if anyone wants to uh get the get one you can say hands-on piece on web app storage and function then you do those two tutorials there in the chat right finish them it will clear your that part of the concept so at the end of it it basically does the resize of the pictures like that there are a few questions in the chat let me just have a look at that uh is event hub responsible in collecting events from a variety of sources like a like centralized repository no not really uh if uh it see if in the repository so event hub is the one where continuous stream is supposed to come like we are imagining that uh uh the the uh centralized repository is not something from where event hub will be getting the trigger but something in this scenario see so let's think of a food delivery application so we have this zumato food delivery application which is pretty popular in this part of the world so let's say that is the application over here and these are the hungry people right now they they are clicking right now right so they are placing the order so order placement right they are placing the order some of them are canceling the order some of them are angry that where is the eta my order was not delivered i want it free of cost right i someone is canceling the order so all the events which are going whenever there are maybe click events imagine click events are there so what whatever data is getting generated we need someone to be and when you when we are working on a very large scale then you really require an injection layer solution to be able to uh ingest that much so this is where event hub is there uh like this any centralized repository explain that scenario further like what are you what is coming to your mind event might not be there so this is let's say event hub here all the orders are getting placed and this is where we have azure functions behind it this these red functions are the order placement so they order they place the order so you want the food they are responsible for placing the order the these purple ones are the order cancellation azure functions so they are canceling the orders so cancel the let's say the green ones are the ones which are uh are handling the the eta requirements that's what they're handling so event hub was the broker in middle like that one more question is there what is the impact of cold start if it incurs message loss now the good thing is uh let's go back to that picture okay so now the good thing is so let me see if i can find that scalable application so this was the picture i was showing you earlier this architecture here now okay let's let's think about this one imagine this uh azure database was serverless over there right this one this was serverless cosmos db can also be serverless just fyi that can also be serverless so imagine they were serverless now what happens if a function didn't got the acknowledgement if the function didn't got the acknowledgement then it will not delete the message from the queue so if there is a cold start and the it took five more seconds for the database to respond function is not deleting anything so we are not getting the data loss it's not like we are just we are not waiting for the response it's basically an acknowledgement based scenario if you don't receive the acknowledgement don't delete the order uh retry try after three minutes try after some xy z minutes like that so retry should be there don't delete anything on first thing now there is one another question in the chat saying which service can we use to store events coming from different sources i am saying centralized repositories now this is where i would just want to ask one little more details over there uh when you talk about uh let's say uh the events which are coming in uh what is the details inside that like uh what is your overall type of scenario like what are the things like where are you getting it from is it like somebody did a merge somebody deleted something notification is it a devops scenario which is coming to mind do let me know about that okay so in the meantime what we'll do is we will just uh go back to the diagram which we we were discussing this thing out okay so one last thing i think we'll be able to talk about service bus i'll just cover this service bus i'll have your questions and i think we'll be on time then so these three things are very important cover these three uh right it's very useful what is service bus let's take this so service bus is the one where uh we have to type of service buses with us we have topics and we have queues right so let's say this is your application a application which is basically the publisher of the data right now or it is the one which is writing the the uh something you can say messages as such this is the service bus here so when you make service bus this service bus will be having uh is having a concept called as queue over there service queues like that so the what service queues will do right so let's say these are b applications over here right so the a applications are basically writing b applications are reading from the queue like this and they both are scaling on their own this is also scaling and the a is also scaling right over here now uh when when this happens so these are the messages here let's use a little less thicker line over there but this is the messages going like that right like this so these are those xml messages like that so the message service bus cues are used whenever you have uh smaller not smaller like your message size is maximum uh 120 256 kb and 1mb so these are the message sizes and servers were bus queue first and first out is totally supported fifo is there like this and this the length of this skew is going to be 80 gb like that but what happens if your applica if if if this is not the scenario you are saying that we actually are not looking for that we have uh we are we are looking for the enterprise service bus like there is a very popular word in the in our world called as enterprise service bus so see now multiple different applications are there we have a application we have b application we have the green color c application as well these green class c and they all are writing into one central service bus over here like this so something like that right so we have a concept called a service plus topic see topic like this over here now check this out so what this service bus topic does and there are multiple consumers also see at the back and you have this d application this is a d application over here and then we have the this uh yellow color uh e is also there and maybe we have a few instances of this one also f over here like that right so they all are consuming now in this scenario what practically you will require is you need some sort of partition so what service bus topic does it internally creates partition like that see these are internal partitions over here so when when somebody is writing the data they write into one partition and when the consumers they consume the data they read from the partition so service bus topic provides that filtering capability it provides those partitioning capabilities so where are a message going right now see all the a message are there in the a partition see like this where are the b messages going right now the b messages are going in this b partition over here where are the c messages going c is writing the message in its own partition like this over here so the rest of the constant will remain same 256 kb message size or 1mb message size first and first out is supported but in the same partition not across partition right in the same partition now when the subscriber subscribes the good thing is they can subscribe to one partition they can also have more than one delivery so if you can see the message coming from a is being read by d e n f right so it can we can have more than one delivery also you don't delete when you read you don't delete you just you do you put a read lock and then don't delete let someone else also read so now different people can subscribe to different messages based on the requirement so this enterprise service bus which if this is your requirement where you want that more than one delivery where you want multiple applications to be using that central message passing mechanism this is when service bus topic is used right so in your dev in your certification you should know how to do the read write how should you write into queues and you should know how to read from the queues how what are the type of locks which are there for the default locks as well how do you extend the lock those type of questions will be there so the the there was a lab which i was giving you which i gave you a little some time back um right over here so you have to perform this one so if i just take you to that one uh you see this asynchronous message passing here uh this one you this is one thing which you have to uh perform like how do you read write from the queues as such right over here apart from this if you want further practice go to ms learn you do further practice from there in ms learn but this is the lab number 10 is what you should do to be able to understand these uh the cues not the other two but the queues as such so let me look at the uh chart window the question is there uh yeah use case is related to devops it's mostly about the storage storing log files which service can be used to store it centrally okay now the okay so check this now if you want to there are two things uh which you can can use uh and so check this so first thing is azure storage account in azure storage account we have one concept called as append so what do we have over here we have append blobs so imagine this is the container like log container let's let's put it like that now this is the log file so this log file goes into up append blob so what is what is uh what what will it be looking like something like this there will be a uri there is going to be a metadata and then we have the object over there so this is the object so this object is actually is having these append you can say append operation is going on over there so uh you can say data is getting appended right now so the logs are supposed to be kept in storage accounts in the append blob over there append blob append blob is designed to keep logs as such of what logs of anything it can be devops logs it can be activity logs error logs any logs can be kept over there but the problem is how will i run queries on top of it storage account is a storage solution so we have one service called as log analytics so what log analytics does log analytics can sit on top of the same storage account like that sits on the same storage account and create the dashboards for you so if if any anything see in azure devops we uh you actually don't require log analytics but uh but if i take you to my devops portal see let me see if i can log into my devops portal and show you in a little bit more about there so this is my devops portal see so basically uh we if you create any release pipeline let me see in which one do i have it i was showing this to somebody last week so i have a few projects there uh i remember making one release pipeline here or in this one okay so now uh if something succeeded or something failed uh the monitoring for that can be done from azure devops as well or what else can i do i can have this storage account configured to keep the logs log analytics on top of it and the dashboard will be created there so this is that central location but the problem is this is not really central it will be in one part of the world like east us uh you uh right like east us west us you have to keep it somewhere so uh this is one of the best location cheapest as well to keep your log files storage account with or without log analytics and see if you can if you don't even require it because in azure devops we basically have the the notification options already already there uh but if you want the analysis on the logs log analytics is the answer for it okay so guys we have around seven minutes left with us let's let's take your questions now i'll uh any question which you have you can ask uh right so let's use that time for this and before that i'll give you the the lab so those who are very interested in what we just discussed see if you do this one case study the probability of you clearing the certification is very high so this is the landing page right this is the landing page where the case study which we discussed is there by microsoft right all the credit goes to microsoft thank them for this now uh the solution is on this page i'll just give the links directly this is the solution over here right over there you need azure subscription without azure subscription you cannot do it uh go for the free trial right and the lab is here like end-to-end deployment uh uh modern application end-to-end deployment is here so the lab is there in two part there is a part a and part b so this is part a and this is part b of the tutorial so over there so those who are interested in the certification go for this see so uh what what we what what is going on uh you create the container registry you create databases event hub is there data factory is coming some extra thing is there don't be shy to learn something something extra iot hub little bit of iot is there a little bit of ml a little bit of azure functions cosmos database uh containers event hub uh then we have the uh the uh the cqrs is there uh change feed is there pilotable of power bi a little bit of this and that so modern modernize your application with data and ai this is the architecture which you are deploying this this workshop will take you four or five hours to deploy but once you do it it will be you will know what you're doing right so this is i'm giving this link in the chat please note it down and use it right it's it goes in the chat note it down deploy that okay let's take the questions now so right so that log logs uh storage account blobs are the one what we are thinking of what is equivalent of windows service in azure what do you mean windows service what is this what do you mean by windows service let me know let me know in case of blobs as well again coming from multiple sources multiple sources a different append so different blobs will be here so imagine this is source number one so this is coming from one source source one this is another log file another append block for the second source so are multiple files for multiple uh you can say sources or every source is writing its own file that's what i will put it like like uh multiple log files are there in the same location like that this is from source number two like this so not the same file multiple files if you if you mean background running process then which what best is an azure web jobs as such now literally uh there are there are two things which you can do uh first is literally called as azure jobs only we have a concept called azure web jobs uh so what happens in the app service plan the job is created and that job runs continuously doing some activity for you second thing is if it is not continuously running you want some scheduled job or you want event driven job then go for azure function so either you use azure jobs literally there's a concept called azure jobs and secondly azure functions those two will be your locations to work with right so like this i'll give the link in the chat this is your how to run web jobs this is the link for it this is how you do it in azure okay so we have few minutes with us guys let me know uh what questions comes to your mind uh the case study note it down uh find me on linkedin as well if you have questions afterwards also feel free to ask uh it made i may respond in few minutes a few hours but i'll definitely respond back sooner or later to you right so let me take you back to that find me on this linkedin page over here and you uh ask your questions later on as well get certified and and if you have questions feel free to ask

Channel: Kasam Shaikh

Views: 3,281

Keywords: How-to-pass-AZ-204, AZ-204-Exam-Preparation, DearAzure, Azure, AzureAI, AzureAiforAll, learntowin, kasamshaikh, MVP, MCT, Azure-INDIA, Clouddeveloper, development

Id: xgewjoY4cD4

Length: 88min 30sec (5310 seconds)

Published: Sat Sep 04 2021