Chris Travers has submitted a new session!

Building Petabyte-Scale Postgres Clusters

At one of the previous places I have worked, we built a semi-structured data store that stored 10PB of log data for the support team's use. This was done with PostgreSQL and involved less hardware than ElasticSearch at a cost of some functionality. Today I am one of the contributors to an open source successor to that project.

This talk will go through the design of the system on a high level, how it works, and how to architect large-scale distributed data platforms using PostgreSQL or other storage technologies.

Open session

These notifications can be turned off on Edit event page.