Ishwar Bhat

I'm a soft-spoken passionate software engineer. I have a total of six plus years of experience. Mainly, I work on web applications as well as delivering them end to end. So, basically, my main technical skills are Python, Django, and I'm comfortable with AWS cloud. And in the front end, I am comfortable with React JS, JavaScript, and the React JS Next.js framework. So, I've worked on various web applications where we have used REST APIs also. We can integrate with the front end. Also, I'm experienced in running scheduled jobs. If in a web application, you need to run Chrome jobs, like that. Also, I'm experienced with end-to-end deployment, like creating CI/CD pipelines, hosting the web application in Elastic Beanstalk, in AWS, and managing all those things. Also, I'm experienced in managing a small team. I have handled a group of developers with a group of 10 developers. So I can lead the team also.

All of asset properties in Postgres. Asset properties, I'm not sure.

Python libraries, commerce. I'm not sure. Maybe Pandas, which will be great for handling Excel's data frames. Apart from that, we can use matplotlib if you want to plot any graphs or anything.

Node Js. Server side Node Js. Mostly, I use only Python in the server side. I have never used Node Js as a web server, but Node Js. Like I said, I am only comfortable with the front end part. So, I noticed we can just create models in Django as we created Python models. We can create models and use an ORM and stuff on top of direct database queries that will help us choose the engines later also. First, we need to use the REST APIs in the server and you like, no HTML editing at all, just the REST APIs, integrate some authentication, the REST APIs. Yeah. Create the app accordingly if we are using Django, as per the use case.

Python framework for server-side logic, we are largest in Django. The thing is it's very flexible, and it's very efficient. You can write logic quickly. You can clear projects very quickly. It provides all the security and all the frameworks for the REST framework, using REST, all the REST API framework. So, if you use Django REST framework, by default, security and all are taken by Django only. Like, X-frame, all those, whatever is there. So, REST APIs, you can easily create with Django REST framework, and you can manage all those settings and all in one place. And, like, if you want more efficiency, Django querying is through query sets. So, one time on next, you run less. You will actually use it. It will be very efficient. So, like, if you want more customization, writing is very easy. But if you want more customization, you can also execute using the Django framework. And, if you want to use it as a server, it's very easy. Scheduling jobs and all is very easy. But if you want to use it in the front end, the Django template model is there again. So, Django templates are very easy to render. That will be very easy.

We'll implement data processing in many ways. A good approach would be event-based triggering. If something happens, we subscribe to that event. And when data changes, we get notified so we can handle the data in Python through various means, such as webhooks or APIs. Apart from event-based triggering, if you want to do it in a more scheduled manner, we can do it in jobs, where we schedule jobs to process the data at regular intervals. There are also other options, which depend on the problem you're trying to solve. Sometimes you might prefer Redis if you want to handle WebSocket-like scenarios. In that case, we can use Redis or RabbitMQ for caching and memory management. Otherwise, we can use SKS if you're dealing with asynchronous tasks and the problem is best solved that way. So there is a task, which I can put in the queue, and my worker can process that task. Depending on the problem, we can proceed with the implementation.

If you want to set up singleton, basically, this is your TV connection, and the connection should happen only once. That's what usually the DB connection should happen only once. Because if you have more connections, it would be like more upload on the server. So, in the init method, assuming we have time to implement, I would make this class instance, my DB connection as a singleton class. Singleton in the sense, only one instance of this class can be implemented or created. So even if you create more instances of this class, it will be only one instance. It will still point to the first instance. So, I'll implement the singleton. There are a lot of ways you want to create a singleton, my preferred way is to override the new method. So, in that new method, we can just pass in the object, and we can make it a singleton. You can use meta classes also. There are a lot of options to create a singleton. Basically, to create a singleton, you can override the new method and easily do that. Then, after that, whenever you create a new instance, it will be the same connection and execute the query using the same connection. Since there can be only one instance, so in this application, there will be only one connection.

Fetch data cannot return the exception as we received it. It's a pure Python. So, basically, in the try catch, I can see the response on JSON. So if it is a JSON response, it's returning just JSON. If it's anything else, it will go into an exception. And if it is going into an exception, we need to return some empty data, some error code or invalid, I should be handling that API error, basically. But if I'm returning a Python exception again, it's not going to work. It's a bad way of doing it. Basically, here, we need to try. And if it is an exception, we need to have some proper formatting. So, my once this response code, I can check whether it's a 400, 500, whatever. Based on that, I need to take action, just not return the exception. That doesn't make sense. At least I can return fetch data as False or something like that or an MDJson response. That might be more appropriate.

Okay, basically, we are deploying a web application, which was developed in Django. So first, we were using normal EC2, but later, like, it was giving a lot of latency, and also we implemented more to elastic means scaling. Then, we used CloudFront to distribute static content, which we stored from S3. So we use AWS S3 for serving static content, making it a public bucket, so static content won't come to the actual EC2 server. After that, we can do a lot of DDoS attacks prevention in AWS CloudFront, so it won't affect our application performance with unnecessary spam requests. So after that, it's easy to improve the creation part by using a lot of debug methods. So, we don't use first-class, last-resort, and Django for database queries. So it will hit the database. Like that, we can do a lot of things. Alright, the DB queries are very efficient. That will also improve the performance. And, like, again, if there is a lot of load, we can definitely implement auto-scaling, so that it will scale up whenever the request is high.

Into enhancing the performance as is, it will totally depend on the scenario. Performance, basically, we use easy-to-use and we will use mostly easy-to-use Elastic Winstock, not directly. So that will help in so few things. If there's something I need to retrieve very often, then there's no need to hit it in the RDS. So I will use a disk cache so that now I will get it quickly. No need to do any connection or anything. If it's only storage kind of thing, we can do it through RDS, like, very easily. But if it's for a temporary purpose, and I just want to get it fast, but it's only for a few minutes or something. I can definitely store it in Redis. Any event-based also, I can store in Redis so that it will announce the performance. And as I mentioned earlier, like, static content, we can serve. No need to go to the server. And