Parallel Hash Index Database Bloat

kuzudb / kuzu

Embeddable property graph database management system built for query speed and scalability. Implements Cypher.

https://kuzudb.com/

MIT License

1.2k stars 85 forks source link

Parallel Hash Index Database Bloat #2625

Open Riolku opened 6 months ago

Riolku commented 6 months ago

After #2615, because we use three disk arrays per index, we use a ton of space for hash indexes on small tables. There are a few opportunities for optimization here, primarily removing the header for the hash index.

Riolku commented 6 months ago

As discussed with Guodong, he will take this on, not me.