Compressing multidimensional data in PostgreSQL
Karl Pietrzak has worked for a variety of companies--large and small--and evangelizes PostgreSQL wherever he goes. His interests include databases and compression, topics that often intersect in interesting ways.
He currently lives in Philadelphia, PA and works for Two Six Labs on a variety of DARPA projects.
Recent growth in the IoT industry has led to increased demands of relational databases such as PostgreSQL. In this presentation, we will use a combination of TimescaleDB and pgpointcloud to:
- create a database solution supports very high ingest rates
- while also providing compression across multiple dimensions
With our dataset (multidimensional radiation/spectral data), we have seen compression rates up to 80% while still maintaining full access of the data from SQL, resulting in massively decreased load on PostgreSQL and massively increased ingestion rates.
- 2019 March 21 15:50 EDT
- 20 min
- Postgres Conference
- Use Cases