Peter Bourgon on CRDTs and State at the Edge

The InfoQ Podcast - A podcast by InfoQ

Categories:

Today on The InfoQ Podcast, Wes Reisz talks with Peter Bourgon. Peter is a distributed system engineer working on Fastly. His area of interest is around coordination free replicated systems. The two engineers talk about the space of Conflict-Free Replicated Data Types (CRDTs) specifically in the context of edge compute. Topics covered on the podcast include Edge compute, CRDTs, CAP Theorem, and challenges around building distributed systems. Why listen to this podcast: - CRDTs (conflict-free replicated data types) are a class of coordination free replication systems (or systems that don’t require strategies such as leader election). - An easy way to think of a CRDT is as an associative, commutative, and idempotent data structure plus the operations needed to use it. The edge is an overloaded term that people tend to define based on where they sit on a spectrum between the customer and the data center. Fastly’s edge is away from the data center and but not to the telephone pole or handset. - RAFT and Gossip are two alternative approaches to using a coordination free replication system like CRDTs. To get the properties of a CRDT and have useful data types, you have to pay a cost in size and often bytes on the wire. These are challenges that continue to need solutions. - Modern Distributed systems and data structures like CRDTs require you to start thinking about state in the system itself. It’s not unusual for a system today to give you multiple results back that the system will have to handle or merge. More on this: Quick scan our curated show notes on InfoQ https://bit.ly/3dTZOO4 You can also subscribe to the InfoQ newsletter to receive weekly updates on the hottest topics from professional software development. bit.ly/24x3IVq Subscribe: www.youtube.com/infoq Like InfoQ on Facebook: bit.ly/2jmlyG8 Follow on Twitter: twitter.com/InfoQ Follow on LinkedIn: www.linkedin.com/company/infoq Check the landing page on InfoQ: https://bit.ly/3dTZOO4

Visit the podcast's native language site