Presented by:

15138544 10210421119420599 1841521026585993990 o

Oleg Bartunov

Postgres Professional

Oleg Bartunov, a major PostgreSQL developer, has been involved in PostgreSQL development since 1996 ( he introduced locale support). He is one of the author of the PostgreSQL extendability infrastructure (GiST, GIN, SP-GiST) and several popular non-atomic data types, such as, arrays, full text search, NoSQL (hstore and jsonb). His latest contribution is a new access method RUM, which greatly improves full text search.

Oleg graduated from the Astronomy department of the Physics department at Lomonosov's Moscow State University. During his professional work at Sternberg Astronomical Institute (SAI MSU) he realized he needed a free and powerful open-source database like PostgreSQL (Postgres95 that time). Since then he extensively used PostgreSQL in his scientific work and many other projects. Oleg is an active member of Russian PostgreSQL community, he advocated the adoption of PostgreSQL by astronomical community. One of his scientific projects make use of scalable indexing of multi-terabytes sky catalogues developed for PostgreSQL, which outperforms known techniques used in commercial databases.

His main interests are Informaton Retrieval, Databases Extensibility, Algorithms and Data Structures, Service-Oriented Architecture for Science and XLDB for Science. In spare time he like to play volleyball, running and trekking in Himalaya.

No video of the event yet, sorry!

TOAST allows to store a long attributes in postgres tuple. Being useful in the past, nowadays it's a sticking point for non-atomic data types like jsonb, which is one of the main driver of growing popularity of postgres. I will present the results of our experiments with modifying TOAST, which allows us to improve jsonb performance by orders of magnitude.

Date:
2022 January 20 15:40 PST
Duration:
20 min
Room:
Market
Conference:
Silicon Valley 2022
Language:
Track:
Dev
Difficulty:
Medium