YoVDO

I Can't Believe It's Not Causal! Scalable Causal Consistency with No Slowdown Cascades

Offered By: USENIX via YouTube

Tags

USENIX Symposium on Networked Systems Design and Implementation (NSDI) Courses Distributed Systems Courses

Course Description

Overview

Explore a groundbreaking conference talk from NSDI '17 that introduces Occult (Observable Causal Consistency Using Lossy Timestamps), the first scalable, geo-replicated data store providing causal consistency without the risk of slowdown cascades. Delve into the design, implementation, and evaluation of this innovative system that supports read/write transactions under PC-PSI, a variant of Parallel Snapshot Isolation. Learn how Occult achieves immunity to slowdown cascades by weakening PSI's replication of transactions committed at the same replica, while still guaranteeing that all transactions read from a causally consistent snapshot of the datastore without requiring coordination in asynchronous replication. Gain insights into the challenges of implementing causal consistency at scale and discover how Occult overcomes these obstacles in this 25-minute presentation by researchers from The University of Texas at Austin, Cornell University, Facebook, and the University of Southern California.

Syllabus

NSDI '17 - I Can’t Believe It’s Not Causal! Scalable Causal Consistency with No Slowdown Cascades


Taught by

USENIX

Related Courses

Advanced Operating Systems
Georgia Institute of Technology via Udacity
High Performance Computing
Georgia Institute of Technology via Udacity
GT - Refresher - Advanced OS
Georgia Institute of Technology via Udacity
Distributed Machine Learning with Apache Spark
University of California, Berkeley via edX
CS125x: Advanced Distributed Machine Learning with Apache Spark
University of California, Berkeley via edX