Presented By O'Reilly and Cloudera
Make Data Work
Feb 17–20, 2015 • San Jose, CA

Maintaining Low Latency while Maximizing Throughput on a single cluster

Yuliya Feldman (Dremio Corporation)
4:50pm–5:30pm Thursday, 02/19/2015
Hadoop Platform
Location: 210 B/F
Average rating: *....
(1.50, 2 ratings)

The good news: Hadoop has a lot of tools. The bad news: Hadoop has a lot of use cases with conflicting sensitivities. And a lot of tools. This talk will showcase how new advances in YARN and Mesos provide the ability to truly run multiple distinct workloads together. This talk is about using SLA and latency rules along with preemption within YARN and dummy resource allocations allow someone to maintain a high cluster throughput while guaranteeing latency SLAs for interactive and low-latency applications such as Apache HBase and Drill.

NOTE: the password to the video is “mapr2015”

Photo of Yuliya Feldman

Yuliya Feldman

Dremio Corporation

Yuliya Feldman is a Principal Software Engineer at MapR. Since joining MapR, Yuliya has worked on a number of products and features starting from MapR admin infrastructure, Map/Reduce framework and most recently on YARN.
Prior to joining MapR, Yuliya worked in diverse areas of satellite image processing, medical robotics, satellite phone communication, e-commerce and Big Data.