Well, unfortunately i am not seeing much difference. I shaved off maybe a second of worst case run.
I guess i should just split the db into smaller ones, since tmpstats are now per-db. Are there any other things i could try? 2017-01-05 8:18 GMT+01:00 marcin kowalski <yoshi...@gmail.com>: > Thanks, i'll redo the benchmarks and report back how things look now. > > 2017-01-04 20:33 GMT+01:00 Pavel Stehule <pavel.steh...@gmail.com>: > >> >>>> > >>>> > This is irrelevant of amount of data restored, i am seeing the same >>>> behavior with just schema restore, as well as with schema+data restores. >>>> > >>>> > If anyone is interested i may upload the schema data + my >>>> benchmarking script with collected whisper data from my test run (i've been >>>> plotting it in grafana via carbon) >>>> > >>>> > Is this a known issue? Can i do anything to improve performance here? >>>> >>> >>> we had 10K and more tables in one database - and we had lot of issues. >>> >>> I know so Tomas fixed some issues, but we need the stat files in tmpfs >>> >>> please, read this article https://blog.pgaddict.com/pos >>> ts/the-two-kinds-of-stats-in-postgresql >>> >> >> http://hacksoclock.blogspot.cz/2014/04/putting-statstempdire >> ctory-on-ramdisk.html >> >> >>> >>> Regards >>> >>> Pavel >>> >>> > >>>> >>>> -- >>>> Jerry Sievers >>>> Postgres DBA/Development Consulting >>>> e: postgres.consult...@comcast.net >>>> p: 312.241.7800 >>>> >>>> >>>> -- >>>> Sent via pgsql-general mailing list (pgsql-general@postgresql.org) >>>> To make changes to your subscription: >>>> http://www.postgresql.org/mailpref/pgsql-general >>>> >>> >>> >> >
