Guten Tag Michael Lewis, am Donnerstag, 31. Dezember 2020 um 18:20 schrieben Sie: > Why is your fillfactor so low?[...] I've just copied what my GUI-tool pgModeler generated as SQL right now, that fill factor might have never been applied at all. > Perhaps a silly question, but do you have an index on captured_on? Yes and that changed a lot some years ago, but doesn't seem to be sufficient anymore. > You included the full DDL for the table it seems so I wanted to ask. No, only the relevant part necessary for my question. How SEQUENCES, PKs etc. are handled is the most important thing right now. > Can you > provide some example queries that you want to speed up? What are you > configuration settings (select * from pg_settings;)? No thanks, I don't want to discuss such things in this thread. I'll asked lot's of other questions regarding my concrete bottlenecks, schema, app logic etc. already and simply want to test with partitions now. That has been suggested in the other questions as well and is what I focus on now. > [...]That is, you could not have a primary key on ID and > partition on captured_on for instance. That's what I understood as well and is the reason why I asked: That means IDs could be duplicated manually within individual partition tables, while I need them to be unique across all of those. Additionally I wonder when IDs are generated by which SEQUENCE etc. Mit freundlichen Grüßen, Thorsten Schöning -- Thorsten Schöning AM-SoFT IT-Service - Bitstore Hameln GmbH E-Mail: Thorsten.Schoening@xxxxxxxxxx Web: http://www.AM-SoFT.de/ Telefon: 05151- 9468-55 Fax: 05151- 9468-88 Mobil: 0178-8 9468-04 Firmensitz: Bitstore IT-Consulting, Frankfurter Allee 285, 10317 Berlin Steuernummer 037/230/30566, HR 27198, Amtsgericht Potsdam Geschäftsführer Janine Galonska
