Abstract The interest and demand for training deep neural networks have been experiencing rapid growth, spanning a wide range of…
CPI2 : CPU performance isolation for shared compute clusters
Abstract Performance isolation is a key challenge in cloud computing. Unfortunately, Linux has few defenses against performance interference in shared…
F4: Facebook's Warm BLOB Storage System
Haystack was the primary storage storage system designed initially for Facebook’s Photos application. Its been around for almost 7 years…
Mesa: Geo-Replicated, Near Real-Time, Scalable Data Warehousing
ABSTRACT Mesa is a highly scalable analytic data warehousing system that stores critical measurement data related to Google’s Internet advertising…
On Designing and Deploying Internet-Scale Services
Abstract The system-to-administrator ratio is commonly used as a rough metric to understand administrative costs in high-scale services. With smaller,…
Coflow: A Networking Abstraction for Cluster Applications
Abstract Cluster computing applications – frameworks like MapReduce and user-facing applications like search platforms have application-level requirements and higher-level abstractions…
From research to practice: experiences engineering a production metadata database for a scale out ï¬le system
Abstract HP’s StoreAll with Express Query is a scalable commercial file archiving product that offers sophisticated file metadata management and…
Replicated Data Consistency Explained Through Baseball
A key feature of all distributed storage systems is their ability to replicate data not just across machines within a…
Apache Hadoop YARN: Yet Another Resource Negotiator
Detailed post to follow… Link to paper
Split Query Processing in Polybase
Abstract This paper presents Polybase, a feature of SQL Server PDW V2 that allows users to manage and query data…