"Modern B-Tree techniques" by Dmitrii Dolgov (Strange Loop 2022)
Vložit
- čas přidán 14. 11. 2022
- B-Trees are probably the most important access structures in databases and file systems, and everyone knows basic ideas about how B-Tree work. But after reading a few white papers about this topic it turns out that these "basics" are about 40 years old. Over the years so many techniques were proposed in order to improve efficiency in one cases, add functionality in others or even embrace new types of hardware.
Inspired by Goetz Graefes and many other authors, we will go through the most interesting of such techniques, see why they are so cool, what kind of use cases they address and how many of them did land in PostgreSQL or other databases.
Dmitrii Dolgov
PostgreSQL contributor, Software Engineer
@erthalion
PostgreSQL contributor, Software Engineer
------- Sponsored by: -------
Stream is the # 1 Chat API for custom messaging apps. Activate your free 30-day trial to explore Stream Chat. gstrm.io/tsl - Věda a technologie
4:00 optimise for read, write, mem consumption, chose 2.
5:00 b tree : read and mem.
6:53 1 index per leaf
Brilliant.
Brillant!
Amazing presentation 🎉
Was wondering on the last bit about data approximation, are there any references to that subject?
B"H" I asked Dmitrii. Here are the references for data approximation he talked about
Galakatos A., Markovitch M., Binnig C., Fonseca R., Kraska T. (2019).
FITing-Tree: A Data-aware Index Structure. 2019
International Conference on Management of Data (SIGMOD ‘19).
doi:10.1145/3299869.3319860
Yes please.
UB-Tree xD