About me

I am a results-oriented cloud engineer in AWS cloud with a background of networking and system administration. In my current role as the Lead Technical Engineer at AT&T labs, I work on AWS Cloud, Linux and Solaris systems, and networking protocols supporting AT&T SDN based DHCP and RADIUS services. During the past few years, I have worked on Docker, Ansible, CloudFormation and Elastic stack using CI/CD tools. I currently hold AWS Professional level DevOps and Solutions Architect certifications and Cisco professional certifications.

Projects

    



Website Provisioning


Repo here


AWS Cloud - S3, CloudFront and Route53

Developed this website from scratch and hosted the code on S3 buckets and served using CloudFront and Route53. Used CloudFormation stack to create S3, CloudFront and Route53 resources and also did a setup of budgets notification to when max threshold and budget parameters are reached.

  • Used Route53 to register a custom domain and a new SSL certificate was created using Amazon Certificate Manager (ACM)
  • Added CNAME records to Route53 DNS Public Hosted Zone to support HTTPS.
  • All the needed resources were created using CloudFormation:
    • S3 bucket with static website hosting configuration
    • CloudFront Distribution.
    • Route53 RecordSet
  • CloudFront distribution was configured to use S3 end point. Route53 DNS HostedZone records were pointed to the CloudFront endpoints.
  • A CloudFormation template is used to setup budgets notification to a specified email with max threshold and budget parameters set.

Automated Deployment


Repo here


Automated Deployment - Docker and CodePipeline

AWS CodePipeline is triggered every time a change is pushed to master branch. It invokes CodeBuild to build the Jekyll website using a jekyll build in a Docker container.

  • Once the static website is built, there was a need for constant development and maintenance of the site resources.
  • To run the website on a different platform for normal development and testing, I chose to use Docker containerization platform and run jekyll in a Docker container. I spun an EC2 instance to be used as a VM and docker is run on the VM to build the website for intermittent manual development and testing purposes.
  • On AWS, CodePipeline and CodeBuild were used to setup a CI/CD pipeline. CodePipeline is triggered every time a change is pushed to master branch. It invokes CodeBuild to build the Jekyll website.
  • The build results are published to S3 bucket for storing the contents of the site.
  • A lambda function will be triggered to invalidate (remove a file from CloudFront edge before it expires) the CloudFront distribution.

Elasticsearch - Log parsing, indexing and visualization


Repo here


Elasticsearch on AWS EC2

Developed elastic stack - Elasticsearch, Logstash, Kibana and Beats and secured with TLS for encrypted communications between application and elastic nodes. Migrated the system to a public cloud using multiple EC2 instances to support each of Elasticsearch, Kibana nodes and Application Load balancer to handle incoming traffic.

  • Installed beats on the application servers to ship the logs to elastic stack nodes.
  • Once the logs are shipped to Logstash, they are matched with regex and extract the needed data while data is indexed and stored using Elasticsearch for faster querying.
  • Kibana is used to display the required metrics in customized GUI.
  • Configured users for secure and modular access.
  • Once the POC was finalized in an on-prem solution, I migrated the system to a cloud solution.
  • Utilized multiple EC2 instances to support each of Elasticsearch, Kibana nodes and used
  • Application Load balancer to balance incoming traffic.
  • Used awslogs instead of Logstash for better integration.

ELK stack Deployment


Repo here


Optimized ELK solution - Continuous Delivery

Implemented ELK stack machines using CloudFormation and AWS ECS to develop in Continuous Integration and Continuous Delivery model.

  • Using AWS ECS, I implemented ELK stack using Docker and created multiple ECS clusters to maintain and serve the incoming traffic.
  • Used CloudFormation templates to build Autoscaling groups that maintain EC2 instances on a per-need basis to serve incoming traffic and also to process constant log data that is shipped from the application servers.
  • This served DevOps methodology by using CloudFormation to track all the changes made to the infrastructure by git and Docker to build everything in a containerized platform.
  • Built a solution where A/B deployment strategy can be utilized for better software delivery.

Python IRC bot


Repo here


Python IRC Bot - text replace

Developed a python Implementation of IRC bot that does sed implementation.

  • Implemented an IRC bot that does a sed like implementation in Python3.
  • While implementing this, I went to all the nooks and crannies to get good knowledge of python and in the process of doing so, implemented this idea.

Education



Master of Science

University at Buffalo

University at Buffalo

UB is amongst the top 50 public universities in the US with a national reputation for excellence and a research-intensive university. As the largest, most comprehensive institution in the 64-campus State University of New York system, the research and people positively impact the world.


Bachelor of Engineering

IIIT-H

International Institute of Information Technology

IIIT-H is one of the top 5 Engineering Universities in India. As an autonomous and not-for-profit university founded in 1998, the institute facilitates interdisciplinary research and a seamless flow of knowledge. Recommended as a premier research institution in India by the US’ National Science Foundation (NSF).