Apache Hadoop attached to Networked Storage


Abstract

Today we provide a reference architecture using Apache Hadoop attached to Networked Storage .vs local storage within servers.

Great care is warranted in designing the storage cluster and network to limit impacts to Hadoop job duration when using Network Attached Storage. Limiting IO latency and providing necessary IO bandwidth are critical to normal, expected cluster scaling and performance.

Why use Network Attached Storage .vs local storage within servers?

Services provided by Network Attached Storage may offer businesses control and security over data that may not be possible using HDFS or another file system. The attached publication provides basic guidelines and best practices for how to size and configure Big Data Networked Storage Solution for Hadoop.

Contents

Chapter 1. Introduction and solution summary
Chapter 2. Solution architecture
Chapter 3. Big Data Networked Storage Solution for Hadoop reference architecture tests
Chapter 4. Conclusions and details of test cases

Download (, Unknown)

Leave a comment