Onkar Rajopadhye

• Worked on Teradata Analytics Platform (TDAP) which is a distributed, shared-nothing analytics platform with advanced analytics capabilities. Most notable among them are: SQL-MapReduce (SQL-MR), SQL-Graph (SQL-GR), SQL on Hadoop (SQL-H).

Projects delivered:

• Monitoring and Viewpoint: Successfully delivered multiple complex projects that continuously monitor TDAP’s Machine Learning Engine’s resource usage such as CPU, Memory, Disk IO and Network IO. Integrated MLE metrics with TDAP Viewpoint UI. I contributed in requirements analysis, design and led implementation of sub-components such as Viewpoint API, Disk IO, dynamically enabling/disabling entire monitoring feature.
• ·Dockerization and Kubernetes: Successfully delivered a set of complex projects to make all Aster services run as Docker containers in Kubernetes. Work involved analyzing latest, cutting edge technologies such as Docker and Kubernetes and their application to existing Aster components/services.
• ·ICE: Served as the primary owner of the data distribution component of Aster Database called Intra Cluster Express (ICE). Work included: preparing a future road map, designing and implementing various enhancements such as: providing network diagnostics in ICE using CRC checksum and timestamp, Redesigning ICE for Docker containerization.

Overall:

• Integrated broad understanding of TDAP product and associated processes and applied detailed knowledge of design interdependencies into internal feature designs. This improved Design quality by aligning them with future architecture.
• Provided feedback and approval to Design reviews, Functional Specification reviews, Code reviews, QA test plan reviews.
• Provided guidance and mentorship various team members and interns.
• Contributed to developing and improving various software development processes to improve developer productivity and work quality.
• Helped foster open, collaborative, friendly work environment by always engaging in constructive dialogue which proved very useful in Agile model.
• Regularly collaborated with multiple external teams such as Strategic Offerings Management (SOM), Viewpoint and Product Architects.

• Shark database is Teradata’s in-house research project for developing next generation In-Memory database as an alternative to Oracle TimesTen and SAP HANA.
• As a member of Database Kernel Engineer in Query Compiler module, worked on design, implementation, unit testing of various enhancements.
• Worked on integrating Teradata's proprietary Kickfire code to run inside PostgreSQL database.
• Implemented various features such as Full Outer Join, Scalable Hash Aggregation, Hash-ordered aggregation and Parallel software query operations.

Yahoo! Everest:

• Everest is a high-performance, scalable, and robust platform for Yahoo!-wide data Warehouse applications and business intelligence (BI) applications.
• Developed various processes like: generating aggregates from given data sets, periodic staging data cleanup, periodic swap space reset, etc.
• Main activities were: Requirement gathering, design, Implement, unit test.
• Daily Monitoring production system reporting and/or fixing problems in co-ordination with Bangalore Operations Team and Santa Clara Team.
• Worked on analyzing, playing around with Apache Hadoop Platform at Yahoo! for migrating existing aggregate generation process from Everest Warehouse to Hadoop.

Kickfire MySQL Analytic Appliance:

• Kickfire is a High performance, appliance based full-fledged database server for MySQL OLAP market. It was ranked #1 database in TPCH audit (2008) for best price/performance ratio.
• Worked as a Developer in Query Compiler Team and implemented various enhancements, made several bug fixes in various Query Compiler areas including Query Execution Engine (QEE), Query Execution Planner (QEP), some parts of Query Rewrite Engine and Optimizer.
• Enhancements: Development of generic sop (gSop) and micro sop (uSop) framework in QEE. Design, implementation and testing of Materialization of shared sub-select in SOPs and NULL handling.

ID-Audit Server:

• ID-Audit Server stores WELF format messages received from Identiforce gateways in a MySQL database. These logged events are used by Identisphere Log Viewer for reporting.
• Implemented enhancements, made several bug fixes and worked on installer related tasks on both Windows and Linux.
• Ported existing code from Windows XP to RHEL4.
• Database fail over enhancement: ID-Audit Server should fail over to slave database (if configured) when master database goes down. We used database replication in MySQL 5.0 to achieve this.
• Installer: Prepared RPM installer for ID-Audit Server and enhanced existing Windows installer.

Oracle Access Manager (Oracle-Oblix):

• Oracle Oblix is an Identity Management (IdM) Server that implements identity management and access control along with workflow management, delegated administration, reporting and auditing.
• Worked in Identity server team and implemented enhancements and bug fixing. Some of which included: (i) Locale Fall back enhancement: Identity server should fall back to default locale (en-us) if a message is not available in current locale. (ii) Ran Purify on Identity Server to find and fix memory related issues.