Parallel Data Laboratory Summer Talk - Chris Carlon

— 1:00pm

Location:
Virtual Presentation - ET - Remote Access - Zoom

Speaker:
CHRIS CARLON , Software Engineer
Google

https://www.linkedin.com/in/christopher-carlon-462b6321/

Rapid Storage: Low Latency Object Storage on Google Cloud

Object storage provides immense scalability, but high latency can make it hard to use for performance-sensitive workloads. This talk introduces Rapid Storage, which brings filesystem-like performance to object workloads on Google Cloud. We'll describe how Rapid Storage moves from traditional stateless requests to a stateful, session-based protocol.

Rapid Storage's protocol amortizes the cost of object storage metadata, enabling low latency data operations. We'll discuss benefits to workloads with dependent reads, like selecting from columnar formats, and how the protocol recovers quickly from transient issues without sacrificing latency.

Rapid Storage also builds on Google systems like Spanner and Colossus to bring appendable objects to Google Cloud. We'll discuss how Rapid Storage provides consistency across multiple write sessions, resulting in a reliable building block for distributed systems.



Chris Carlon is a software engineer at Google, working on distributed storage systems. Before Rapid Storage, Chris was a software engineer at Jane Street, and before that he helped design and launch Google's AlloyDB. Chris graduated from Duke University with a BS in Computer Science/BA in Mathematics. 

Zoom Participation.  See announcement. 
 

For More Information:
karenl@andrew.cmu.edu


Add event to Google
Add event to iCal