Sunil Mallesara

Hi. Yeah. Uh, this is Sunil Ghor. Uh, I'm a principal engineer in Australian Stock Exchange working remotely in Bangalore. I have around 15 years of experience, Uh, entirely in Java domain. I've been working as Java developer right from the beginning of my career. Right now, I'm working as a principal engineer with, My day to day activities involving system design and requirement analysis and coding and code reviews and, Uh, response responsible for team management. Uh, typical day, uh, starts with the SCUM meeting wherein which we discuss about The day to day progress blog posts and, um, how, uh, actively we can engage with each other and decide on what should be Prioritized and work on the same. This is my, uh, current profile, uh, at a very brief Yeah.

Okay. Uh, to design a scalable and, uh, secure, uh, API endpoint, uh, we need to start with, identifying what kind of, uh, data that it is processing. And, uh, depending on, uh, the level of A processing that it needs, we can identify, uh, the microservices that it needs to, uh, have. So if it can be, uh, distributed among different microservices, we can scale up using our cloud infrastructure these days. That would be one of the approach that, uh, I would prefer for addressing the scalability of, uh, of any application. And, uh, when coming to the secure API endpoint, uh, we can explore using token certifications in those lines Or to make the API available only for the authorized users or the applications.

Okay. If we are converting 1 with Java application into a microservice architecture on Kubernetes, very first thing is to understand, uh, what this application Java application is doing on that application, we have to understand how many different kind of operations that it is doing depending on which we can actually segregate interdepend independent set of operations and group them into 1 microservice And then, uh, I'll make them deployable on, um, our airport, uh, on Kubernetes. So, uh, this is how, uh, how we can start, uh, changing the design of our application. So this would be the very first step. And and then coming on to the, uh, uh, Kubernetes architecture, uh, we can always think about, uh, uh, deploying on port and, uh, having the customer are to manage, uh, this port and, uh, uh, the backup strategy and failure strategy, I think, can be defined on the Kubernetes. So this would be, uh, my approach to, uh, start thinking on, uh, converting a monolith, uh, Java application into a Microsoft architecture on Kubernetes.

Okay. Um, if we have to design a scalable real time data pipeline, AWS, How we can, uh, use existing capabilities of AWS are to address our issues For minimal latency and data loss, uh, we had to go with the best practices on AWS and the available features, uh, to define, uh, our strategy. Uh, I'm not able to give, uh, the exact solution at this stage, but, uh, I would start with, uh, read reading through the edit, uh, AWS capabilities on this front, uh, to address, uh, data

Okay. If we are migrating, uh, refactoring the last code base, uh, code base from Java to Scala, I would start from scratch. Uh, take one thing at a time and identify the services which can be taken, uh, in, uh, independent of the remaining part of the codebase. So convert, um, our refactor, uh, once at a certain time, uh, and, uh, upgrade each of these So, uh, stage by stage, make sure they're dependent on, uh, areas are upgraded together Uh, so that, uh, there won't be any much, uh, downtime, uh, in terms of upgrading these services uh

Yep. Again, uh, coming to the AWS. I have, uh, limited working experience with AWS. Uh, now I have a very good experience with GCP. So, uh, to answer questions for AWS, I would definitely start looking at the similar options uh, that we have and then identify, uh, the corresponding solution on the AWS. There is always, uh, a lot a lot in which I'll, uh

To ensure high availability and disaster recovery for a distributed application on AWS, we have many features which Make sure, um, uh, the higher availability. We have the availability zones, uh, which will address, uh, high level availability. If some, Some of the zones are down for some technical issues or any any issues for that micro. Uh, the surface on the other availability zone would Back it up and then make sure, uh, there is no performance impact or, um, downtime. And in case of disaster recovery as Well, uh, if there is some region, uh, has some, uh, uh, major disasters, it could be as measured as, uh, earthquakes. So, uh, now the the remaining, uh, available day zones will, uh, pick it up, uh, pick the request up and then scale up to, Uh, address the uh

We are assigning null, uh, interpretive data type Paint. Uh, I think we can go with, uh, the default value of 0 instead of, uh, assigning it now.

To To monitor and auto scale Java application within the Kubernetes ecosystem, I would use the cluster management to manage uh

To optimize JV's JVM performance for a Java application in a containerized environment. Um, I would, uh, look for the memory consumption, uh, from the application, uh, uh, during the application runtime. So if it's using a a heavy Number of objects at the same time, I would, uh, look at optimizing, uh, the load By, uh, uh, looking at the parallel executions, uh, if there is no other way, we'll have to increase the memory. Otherwise, We can always refactor the code to make sure, um, what can be executed, what can be sequentialized uh

AWS provides, uh, CloudWatch to, uh, track the application health. We can actually integrate into The AWS services to make sure there are relevant alerts and backup, uh, steps that we can take, uh, in case of any application, uh, going down. So AWS provides robust Features, uh, which can be leveraged, uh, to manage, uh, our applications and and their

I'm uh