There is parallel sequence scanning coming in 9.6 -- http://rhaas.blogspot.com/2015/11/parallel-sequential-scan-is-committed.html
And there is the GPU extension - https://wiki.postgresql.org/wiki/PGStrom If those aren't options, you'll want your table as much in memory as possible so your scan doesn't have to to go disk. On Thu, Mar 17, 2016 at 5:57 AM, Artem Tomyuk <ad...@leboutique.com> wrote: > Hi All! > > Is Postgres use shared_buffers during seq_scan? > In what way i can optimize seq_scan on big tables? > > Thanks! >
