A Rust Framework to Simplify Distributed Programs
Constructing software program that runs throughout many machines is notoriously troublesome. Builders need to grapple with issues corresponding to race situations, partial failures, and message ordering. Notably, one class of distributed software program has largely escaped these burdens. A distributed database can unfold a single question throughout 1000’s of machines, dealing with the coordination, failure restoration, and ordering internally. This raises a pure query of why general-purpose distributed programming can’t really feel the identical means.
It is a extremely sensible downside at AWS, as a result of the reliability of cloud infrastructure is dependent upon getting distributed methods proper at huge scale. Joe Hellerstein spent thirty years as a database and distributed methods researcher at UC Berkeley, the place he pioneered a lot of the foundational pondering on making use of database concepts to distributed programming. He’s now at AWS, the place he works to carry his analysis into manufacturing by way of Hydro, which is a Rust framework to carry declarative queries to general-purpose distributed programming.
On this episode, Joe joins Sean Falconer to debate how concepts from the database world might make distributed programming dramatically less complicated and safer.
Sponsorship inquiries:
[email protected]