News for hackers. Stuff that matters.

A comment on PostgreSQL B-Tree index deduplication. Parent comment

Re:PostgreSQL B-Tree index deduplication by Skinney on Sunday September 06, 2020 @10:53PM (#24396253)
Roughly speaking: smaller index size = greater performance.

The index also has to be read from disk (if not already in memory), which is faster the smaller it is. But also, once in memory, smaller size makes key-lookups or traversals faster.

How much faster? It really depends on your index.

[ Parent | on HN ]
Re:PostgreSQL B-Tree index deduplication by petergeoghegan on Sunday September 06, 2020 @11:25PM (#24396376)
B-Tree deduplication can also help to control duplicate index tuples caused by MVCC "version churn" -- including in unique indexes. See the paragraph that begins with "B-Tree indexes are not directly aware that under MVCC..." here: https://www.postgresql.org/docs/devel/btree-implementation.h...
[ Parent | on HN ]