Instructors will introduce the Platform MapReduce architecture, enterprise-required features such as High Availability support, as well as provide an overview of Apache Hadoop and MapReduce. Discussions will cover development and IT topics including integration and advanced management of Hadoop programs.

The lecture is approximately 2 hours in length, and will be offered twice on 12th of October, from 09:00 – 11:00 and 14:00 – 16:00. A question and answer session will immediately follow each lecture. Refreshments and sandwiches will be provided between 11:00 – 14:00

Programme

• What is Platform MapReduce
• Common pain points in the market

Review of Hadoop and MapReduce (20 minutes)

The Architecture of Platform MapReduce (20 Minutes)
• Key features which address common pain points
• Resource Management / Consumer Setup for MapReduce Applications (multiple job trackers)
• Installation Requirements / Specifications
• Data support – File systems & Databases

What is in it for developers and administrators? (50 Minutes)

• Application Development – Platform MapReduce APIs
• Converting an existing Hadoop application
• Direct API (show) and Adapter Logic (Explain)
• Job Execution Discussion – Single Job, Multiple Jobs
• Advanced job prioritization and scheduling (Architectural explanation of the features)
• Setting Prioritization
• Fair Share
• Pre-Emptive
• Threshold Based Allocation
• Resource Reclaim
• Resource Blocking
• Resource Draining
• Guaranteed (owned) Resource Scheduling

What is in it for developers and administrators? (15 Minutes)

• High Availability – Discussion
• Job and task recovery logic
• HDFS NameNode failover logic

• Performance Discussion
• Wordcount, TeraSort, Multiple job execution – Compared to Hadoop.
• Troubleshooting / Debugging Discussion
• How to add a second Job Tracker (MR Application) to the same set of resources.
• How jobs are executed across two Job Trackers while leveraging the same set of resources.

Wrap Up: (5 Minutes)

• FAQ Review
• Where you can download more information.
• eMail alias for technical questions

Questions and Answers - Immediately following

ABOUT SIMON WATERER

Simon Waterer is a Senior Solutions Architect with Platform Computing, a leading provider of HPC software. More about Simon Waterer