| From: | Justin Pryzby <pryzby(at)telsasoft(dot)com> |
|---|---|
| To: | Alvaro Herrera <alvherre(at)alvh(dot)no-ip(dot)org> |
| Cc: | Peter Eisentraut <peter(dot)eisentraut(at)enterprisedb(dot)com>, Amit Langote <amitlangote09(at)gmail(dot)com>, Japin Li <japinli(at)hotmail(dot)com>, Zhihong Yu <zyu(at)yugabyte(dot)com>, Simon Riggs <simon(dot)riggs(at)enterprisedb(dot)com>, pgsql-hackers(at)lists(dot)postgresql(dot)org, Tomas Vondra <tomas(dot)vondra(at)enterprisedb(dot)com>, Daniel Westermann <dwe(at)dbi-services(dot)com>, Erik Rijkers <er(at)xs4all(dot)nl>, Jaime Casanova <jcasanov(at)systemguards(dot)com(dot)ec>, Andres Freund <andres(at)anarazel(dot)de> |
| Subject: | Should use MERGE use BulkInsertState ? |
| Date: | 2022-05-11 12:51:48 |
| Message-ID: | 20220511125148.GK19626@telsasoft.com |
| Views: | Raw Message | Whole Thread | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-hackers |
Should it use it ?
It occured to me to ask when reading Bruce's release notes, which say:
| [MERGE] is similar to INSERT ... ON CONFLICT but more batch-oriented.
Currently, INSERT *never* uses bistate - even INSERT SELECT.
INSERTing 10k tuples will dirty 10k buffers - not limited to the size of a
strategy/ring buffer. Currently, MERGE will do the same.
I had a patch for INSERT last year.
https://commitfest.postgresql.org/35/2553/
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Masahiko Sawada | 2022-05-11 12:55:00 | Re: Support logical replication of DDLs |
| Previous Message | Matthias van de Meent | 2022-05-11 12:19:20 | Re: First draft of the PG 15 release notes |