Souvik Sarkhel

Seasoned DevOps and SRE professional with 10+ years of experience driving reliability, automation, and scalability

across large-scale platforms like Disney+ Hotstar. Proven track record of building end-to-end CI/CD pipelines

(Jenkins, Spinnaker, Istio), automating infrastructure with Terraform, and integrating DevSecOps practices. Expert

in observability solutions using ELK, Grafana, CloudWatch, and custom telemetry tools to enhance system reliability

and incident response. Deep expertise in Big Data ecosystems (HDP, CDH, Ambari, Cloudera Manager), including

deployment automation and marketplace product development. Adept at optimizing operational efficiency, disaster

recovery planning, and performance tuning in hybrid cloud and containerized environments.

Total Experience: 10 years 5 months

• SRE:
  - Automated Observability and monitoring stack and setup Alerting along with Incident management process to reduce MTTD and MTTR.
  - Event processing pipeline using Flink to process realtime events and store in a TSDB for alerting using Anomaly Detection.
  - Design and implement a system that can be able to centralise the alert definition.
• Capacity Planning for AWS Managed ELK:
  - Determined instance types, number of Datanodes, and Master nodes.
  - Calculated required disk space and shards for a PROD ELK cluster.
  - Implemented index lifecycle management based on performance team inputs.
• AWS Infrastructure DR:
  - Designed zero downtime switchover plan for AWS services in case of regional outage.
  - Utilized Terraform, AWS CLI, and Jenkins for automation and implementation.
• Jenkins DevSecOps Pipeline:
  - Implemented static code analysis using SonarQube.
  - Conducted container vulnerability scans with Stackrox.
  - Ensured source code security with Veracode and Blackduck scans.
  - Conducted REST API testing using Burp and automated Jira ticket creation on failure.
• Jenkins as a Service on Kubernetes:
  - Deployed Jenkins on demand in Kubernetes with EBS storage.
  - Enabled dynamic agent launching as needed.
  - Executed service tests on Kubernetes via Spinnaker pipeline.
• CD Phase:
  - Created complex Spinnaker pipelines using Dinghyfile.
  - Orchestrated landscape deployments with Spinnaker and Terraform.
  - Utilized Helm3 for microservices deployment.
  - Implemented Canary Deployments and Istio custom gateway deployment.
• CI Phase:
  - Developed Jenkins pipelines as code and shared libraries.
  - Monitored Jenkins node metrics and created alerts.
  - Visualized build telemetry using ELK stack.
• Infrastructure Automation using Terraform and Monitoring/Visualization:
  - Developed Terraform modules for deploying AWS resources.
  - Set up Cloudwatch alerts and dashboards in Kibana.
  - Ensured monitoring and visualization with Cloudwatch and Grafana.
• Solving R&D Issues:
  - Improved Metadata Scanner performance using OrientDb Graph.
  - Enhanced API performance with multi-threading in Java.
  - Developed Java API with Spring for product operations.
• Kerberos Setup in Kubernetes:
  - Developed Docker images for KDC master-slave setup in Kubernetes.
• Custom Service for CDH Stack:
  - Developed Spark Thriftserver service for CDH with Kerberos Authentication (https://github.com/cloudera/cm_csds/pull/5)
  - Handled performance tuning in Spark ThriftServer.
• Datalake Development:
  - Deployed CDH distribution on bare metal and containerized environments.
  - Implemented Kerberization and service discovery.
  - Developed automated test scripts using Python.
• Custom Stack for Apache Ambari:
  - Installed and distributed services in a cluster with custom services in Apache Ambari.
  - Implemented continuous monitoring and one-click installation with Chef.
• Big Data Stack on AWS:
  - Created AWS resources and managed Hadoop clusters using CloudFormation and Chef.
  - Developed end-to-end solutions for deploying big data stacks on AWS Marketplace and TestDrive.

• Achieved 50% cost savings for metrics collection using OpenTelemetry Collector.
• Achieved 20% decrease in incident volume and continuous improvement within the Site Reliability Engineering SRE) domain by automating, refining internal processes, conducting regular reviews.
• Reduced MTTR by 33% by deploying automated alerting systems and refining communication protocols, ensuring rapid issue detection and resolution.
• Reduced deployment failure by 25% by using automated gates in Jenkins, E2E tests and Canary verification.