L V Rao Jakka

I have been in the IT industry for more than three years. I first started my career with solutions, and I have been in the field for more than half a decade. I work for a product called Harishta, which is one of the biggest technologies in North Korea, currently based in the United States. I checked all the possible options. And, I have done some phone jobs to manage all the Linux servers and their configurations. At the time, I left the software solutions as the lead of the team. I used to hand over the project from development to production. I have also developed software in both cases, including fresh food ordering. Throughout my entire career, I have worked completely normally. However, I had a word to say. Everything is going well. I'm currently working on a part called Huddl Retreats, which is a research team. That part is almost completed and ready to deliver. In this part, I will use Stripe for payments and a regular model in MongoDB as the backend. You mentioned that the front end is open to React, handled by another team. Do you know any future projects? I need to, Google the other company, and I'll refer to the manager.

So, generally, to diagnose any performance issues, there were some key points. It depends on what database we are using and how frequently I mean, how many queries we are making. The most final thing we were missing, we are looking at Atlas. So we'll definitely use the performance adviser to check in, which will suggest all the required indexes and the other key points to boost the performance. Other than that, we'll check any synchronous process, even the way which is blocking. The third point that might also be a good reason to use the cluster package. So we can get multiple worker modes, which will handle all the incoming reports.

The structure indicates.

So again, regarding the notice of the issue, how can you do in any platform for integrating error testing? I had an experience using the Apache Kafka and Huddl. So we would publish messages after they were processed. Though they would be slower and at the same time, when there were particular errors, we would send an email to a particular team which would handle this kind of errors.

So Node. Js, the patterns we used to follow, all of the things. One quick release, what we file. Yeah. We're all what's our role? Looking to put at that time, for transition, we can use the MongoDB transactions.

Structure actually, yeah, organizing unstructured language, but we can differ with mongoose. And we can store all the models at one place and some routes at one place. Generally similar to an MVC framework, but without views. Yeah. That is the best way to structure a project using a note. I'm going to

Important, I mean, the major other features we can include, like, we can put a try catch and we can do error handling first, by putting not. So we will avoid gate missing of the reference conditions.

So that this will have pretty function. If they're a callback, that is very functional. So that will function for each data, which will accept a functional increment and it makes a database query. And using the same calling the fixed rate of function and passing it as an argument, which will return the results depending on the query result.

So handling notices that can handle millions of confirmed database operations. We totally depend on it. If you want to add multiple API requests, it's a different story. We need to have multiple workers, for instance, and we need to have a cluster. I mean, we'll set up notes. So they all can be able to serve, and they have to be diverse, and they can be defined with notes. The handling of DNS, but when it comes to making concurrent operations on MongoDB, we can use the fact that not all operations will take up the area of promises. This is a new area of operations. What we want to perform on MongoDB, and all the operations will perform that.

So manage multiple configurations. We generally use a dotenv file and a JSON file. But when it comes particularly to Edge, we can pretty much use that keyword.

There was a time when I had to implement a feature in Projis. Generally, as I said, we have managed to find the hashtags. And the hashtags weekly have promotional offers. So every Wednesday, we get a huge load on the server. It was very huge. So we have tried to keep our daily place in the office. Then we found out there's one place. We are creating a hash map to call it output to make a request. And for the hash map, we are using a crypto. Crypto is kind of a blocking issue at some point. So, we have figured it out. But we don't have a choice. We have to use it. But there is a major issue which is causing the problem. So for this big issue, I solved it using gRPC. I created a gRPC service and I created a RPC call to generate a hash map. I connected my code to the gRPC service and I called it. I'm calling that RPC function. What it did, it created another instance, which is a blocking kind, but my actual API is not blocked. It used to call a function in a regular operation, in simple words. It used to go more than 100% of CPU in all regions. But after implementing this, it came out at 40 to 30% on Wednesday.