Lerin muthuraj

Do you help me understand more about your background by giving brief introduction of yourself? okay. So, my name is Lauren. I am a DevOps engineer and a cloud solution architect, who a solution provider for any cloud solution needed for many on premise to cloud migration stuff. So that area is of my core specialty and, involving into CACD progress of any infrastructure and architecture, and maintaining databases and providing application support to troubleshooting, and maintaining, like, a highly available and a highly scalable environments, for the client, it's my sole, like, my main duty here, as my as a engineer. Yeah. That's it.

Secure AWS. Okay. to outline, a secure AWS architecture for new application. Right? okay. That, there are multiple ways, to create a secure architecture. So the first one, right, is to go in a public and private subnet. We should have both public and private subnet. in the private subnet, we should have the, servers and servers or any Kubernetes clusters. And whatever thing that we are hosting, the application should, be in a private subnet, and the public subnet should have should have a bastion instance or a or the load balancer and all the other stuff should be in the public subnet. it should have an application load balancer as well as a network load balancer to use a WAF rules. And we can also use, CloudFront, to provide additional support from DDoS attacks and other stuffs. so, yeah, this is the, main architecture that I can use, the main thing is to use public and private subnet concepts, so that the private subnet will be accessible only to the, specific people who have access to it, and it will be exposed to the world only through the load balancer. Yeah. That's

In what scenarios would you suggest using Terraform instead of, AWS CloudFormation? Okay. TerraForm is a tool. Right? CloudFormation will be used only for, AWS related, resources. So, in any case, we are using, some other resources like, RabbitMQ or other things, like, apart from AWS that we needed for our projects. Right? So that may be some dependencies, that, we need to do an infrastructure that may not be, inside the AWS provider services. So for that, I will go along with the Terraform. So Terraform has more than around 100 service providers, 100 providers. so, we can use that. Right? It has provided for Kubernetes as well and, OpenSearch, Grafana, and everything that are from us. So yeah.

Set up monitoring for an application deployed in Kubernetes on AWS. Okay. application monitoring for Kubernetes can be done 2 ways. We can use the Kubernetes dashboards, to we can yeah. We I can install Kubernetes dashboard to look at the metrics, and I can also install a CloudWatch agent in my Kubernetes cluster. So that will give a detailed analysis of the of the metrics of the parts and what CPU utilization other parts are taking, and I can Also, direct the logs to the CloudWatch. And, the part logs, I can take it to the CloudWatch, and I check if there is any part failures. I don't know what are the By checking the logs, I can get to know how the part is how the part got failed. So why the why the reason the part is getting restarted again and again?

Do you ensure the back end compatibility of a Java application in CACD process? And a new version is getting Okay. How did you back end compatibility of a Java location. Okay. I'm not sure how to do the patent compatibility check, but, what I would do is right? so when a new version of Java that we are using is being released, right, so that it needs to be supported Click, while starting the application, right, it needs to be started without any errors. And the API calls that we sent to the back end, right, it should provide a Proper response we can use. Once the c the CD process is completed, we can start the site and then use postman or some Postman to check the API calls. If the API calls is are properly working, the front end is properly connecting to the back end or not, we can ensure this one, and we can also integrate SonarQube and other stuff to check if the proper versioning is there within the front end and the back end

Your experience as implementing high level system design with an emphasis on lean agile principles, certain less downtime or fewer production errors. Okay. Okay. 1st of all, right, in agile methods, once, we the production errors. Right? We will be already neglecting. We'll be eradicating these production errors in the staging environment itself. So while moving into the production, we can, minimize the errors in such a way and, for downtimes in production environment. Right? To maintain less downtime, we should have a highly available and high okay. Sorry. Highly elastic environment, which has multiple servers in multiple regions under load balances, which has an, what can I say, automatic failovers? So once one server is getting failed over, we can move it to that where we can move the reroute the traffic to another server. So that is the best solution for maintaining less downtime or fewer production errors.

Given the below Java code, you received a null pointer exception at runtime. Find the possible root cause. Public class test. Private static string is Public static wide mainstream. Yeah. Okay. Okay. They have created a string, public static wide string as argument system dot out. Print line. okay. But they have not passed any variables or, they have not set the starting point for the string that may cause an

Consider the following Kubernetes YAML configuration for deployment. Stop the bug that will prevent successful implementation. Stop. Replica smudge labels, app, engineering's template. Okay. That is a labeler. You have mentioned an NGINX and an NGINX one label. the label should be similar. we cannot mention 2 different tables in an

How will you handle version control and deployment in multi cloud end to ensure consistency across production and nonproduction setups. Okay. Version control and deployments in multi cloud environments. Okay. so for version control, right, where for version control we'll be having in Git. Right? any branching strategies, we'll be using proper branching strategies across multi cloud environments. We'll have 1 GitLab repository, that we can use to store the code with. And, for multi cloud deployments, right, we need to use an CACD architecture, and we need to maintain centralized repositories for the images. So from which, so once, from the from the code repository, right, from version control. We'll be getting the code, and we should have the Dockerfile or something, how to build the coding. I'm talking in case of Kubernetes. There might be monolithic too for Kubernetes while building the image. Right? We should have the docker file inside the version control. so this Jenkins, right, will build the image and push it to a centralized repository, which repository this repository 3 should be used by the multi cloud architecture. So by this, right, this one image will be shared to the entire multi cloud, whatever cloud maybe. It's Azure or AWS, whatever, EKS or AKS or, GKE, whatever it may be. they will fetch the image from the same containers, repository. So the version will be handled same across, all the environments in all the multi clouds, And similar goes for the lower environments as well staging, the testing or development server.

What steps do you take to perform root cause analysis when your application? How will we prevent it from happening in the future? Okay. the basic one the basic thing for us to do to troubleshoot the root cause is to check the logs, for any errors so for any errors or exceptions that has been passed to the log side, we can check it from any monitoring tool, Or we can take it from the application logs or the pod logs, the container logs. So that will help us to understand if there is any, What kind of error is that? Is it is it an application level error? Any functional, like, is there any functionality error? Or is there any application downtime error or any network error? We can separate the errors in this way, and then we can Proceed with the troubleshooting. If there is any network error, we can check from the networks. Say, if the part is not exposed correctly, or is there any, anything wrong with the load balance? Or is there any That certificate, renewal need to be done. from the application side, if the API call is not getting properly, if any put target request is not being posted properly, We can check that too. And, in case of any deployment failures, if there is any file missing or, the image pool might be missing, the part may the part may be failing to pull the image from the container. So this, we can check from the