Meetup @ Yahoo - Awesome talks from Uber, eBay, Cloudera, Verizon Media

******
Please register via Eventbrite:
https://www.eventbrite.com/e/hadoop-meetup-yahoo-talks-from-uber-ebay-cloudera-verizon-media-tickets-78705479207
******

Join us at Yahoo’s HQ for awesome presentations (Uber, eBay, Cloudera, Yahoo/Verizon Media), conversations, & networking! Pizza & refreshments will be served!

[Location & Parking]

Yahoo Campus, 701 1st Ave, Sunnyvale (Building C, Classroom 4)

Please park in the garage attached to Building C, on the 3rd floor.

[Agenda]

5 - 5:45
Pizza, cookies, refreshments, & networking

5:45 - 6
Welcome & Intros

6 - 6:45
Raising the performance bar for stream processing with Apache Storm 2.0
Roshan Naik, Lead - Real-time Compute Platform, Uber

The effort to rearchitect Storm's core engine was born from the observation that there exists a significant gap between hardware capabilities and the performance of the best streaming engines. In this talk, we’ll take a look at the performance and architecture of Storm's new engine which features a leaner threading model, a lock free messaging subsystem and a new ultra-lightweight Back Pressure model.

6:45 - 7:15
Quick Intro to Maha: Open source framework for rapid reporting API development; with out of the box support for high cardinality dimension lookups with Druid
Pranav Bhole, Sr Software Engineer, Verizon Media

7:15 - 7:45
HDFS Cluster Optimization in eBay
Yiqun Lin, Hadoop Team, eBay + Apache Hadoop Committer / PMC member

On eBay, we have many large HDFS clusters with thousands of nodes. We face many stability/data availability problems in our cluster. Today we want to share some optimizations we did in the system layer or HDFS level to improve our clusters. Besides, that makes our cluster more stable than before.

7:45 - 8:15
Ozone - Object Storage for Big Data
Arpit Agarwal, Senior Engineering Manager - Storage Team, Cloudera

Ozone is an Object Store for big data that is designed to keep the best parts of HDFS while scaling to billions of files. Ozone is designed to support the Hadoop ecosystem with applications like MapReduce, Hive, Spark, and Impala working out of the box. This talk gives an overview of the Ozone architecture and describes how we approached solving some of the scale limitations of HDFS. We will also look at the current state and future roadmap.

8:15 - 8:35
Storm 2.0 - Features and Performance Enhancements
Kishor Patil, Principal Software Engineer, Verizon Media + Apache Storm PMC

Please register via Eventbrite:
https://www.eventbrite.com/e/hadoop-meetup-yahoo-talks-from-uber-ebay-cloudera-verizon-media-tickets-78705479207

• • • *

[Speaker Bios]

• Roshan Naik, Lead - Real-time Compute Platform, Uber
  Architect of Storm 2.0's new high-performance execution engine, the Kappa+ architecture and Hive's transactional streaming ingest APIs. He is a committer on Flume, Streamline, and Storm. Roshan is a technical lead for Uber's streaming analytics platform.

• Pranav Bhole, Sr Software Engineer, Verizon Media
  On the Verizon Media for Ads Reporting team. Pranav works mostly on backend APIs and data pipelines responsible for serving low latency APIs powered by big data engines like Druid, Presto, Oracle, Hive etc. He is a founding member and major contributor in the Open Source Project Maha (https://github.com/yahoo/maha).

• Yiqun Lin, Hadoop Team, eBay
  Apache Hadoop Committer / PMC member, focused on Hadoop HDFS module and makes many contributions to the HDFS RBF/Ozone. 200+ patch contributions.

• Arpit Agarwal, Senior Engineering Manager - Storage Team, Cloudera
  Apache Hadoop PMC member and committer since 2013.

• Kishor Patil, Principal Software Engineer & Apache Storm PMC, Verizon Media
  Apache Storm Committer and PMC member. Apache Spark contributor.

Please register via Eventbrite:
https://www.eventbrite.com/e/hadoop-meetup-yahoo-talks-from-uber-ebay-cloudera-verizon-media-tickets-78705479207