Madhu Nitesh Bandi

Software Engineer – II

FactSet Research Systems•  June 2022 - Present•  Hyderabad

Working on the project of building a standardized, reliable, and scalable observability platform intended to accept log messages and metrics from many disparate sources and persist them to a unified data store. • Designed and implemented background (CRON) jobs/tasks using Python, Celery, Redis, RedBeat Scheduler, and Flower to curate Elasticsearch clusters, calculate health metrics, and update cluster statistics. • Developed a web application and REST APIs to provide a user interface for stakeholders to interact with our platform. Utilized the Proxy design pattern to create a common proxy Elasticsearch service, enabling a single instance of Kibana to read from multiple Elasticsearch clusters and to serve our search APIs. • Developed an AWS Lambda function that calculates heartbeat lag and uploads it to CloudWatch. Configured CloudWatch alarms based on the heartbeat-lag metric to automate the autoscaling of Logstash indexers in case of increased load, reducing the operational burden on the on-call person. • Created a Terraform module to provision an AWS Lambda function and an ECR repository and developed a unified Terraform module by abstracting three different modules. • Migrated deployment pipelines for web and background-task applications from Jenkins to GitHub Actions. Additionally, implemented a deployment pipeline for a GitHub repository that builds a Docker image and pushes it to Amazon ECR. • Converted stored procedures from MSSQL Server to PostgreSQL and refactored Node.js APIs to ensure compatibility with PostgreSQL. • Working on migrating existing IAAS Elasticsearch clusters to AWS OpenSearch clusters to leverage the benefits of ultra-warm and cold storage for cost optimization.