On Friday, February 7, 2025, me nefcanto <
sn.1361@gmail.com> wrote:
Hi, thank you for the response. If we analyze semantically, it had to be on_type_error or something. But what matters is the problem at hand. Inserting a million records not in an all-or-fail is a requirement. What options do we have for that?
In core, nothing really. As you see we are just now adding this kind of thing piece-by-piece. Insert has the “no duplicates” solved but not other issues, while copy solves for malformed data at present. The generality of the name is because future errors to be ignored can then be added without having an option for everything single one of them.
You might consider searching the internet for solutions to this long-standing need.
The one project that comes to mind, but I’ve never used, is pg_bulkload.
In any case, this is now a discussion better suited for the -general mailing list.
David J.
