Sorry, I made a mistake at the moment to copy & paste the sequences.
But this finally the problems was the filter to do dump, when I use --exclude-table option with pattern projects_historydata_* for some reason the sequences was excluded finally I used the pattern projects_historydata_[0-9]* to exclude only the tables that I want to exclude.
I don´t know if that behavior is right.
Regards,
Jaime Acevedo E.
El jue, 13 may 2021 a las 14:11, Tom Lane (<tgl@sss.pgh.pa.us>) escribió:
Alvaro Herrera <alvherre@alvh.no-ip.org> writes: > On 2021-May-13, PG Bug reporting form wrote: >> I'm using pg_dump to dump my database version 12.5 (RDS), pg_dump version >> 12.6 but I have a the following table
> How did you get this table definition?
Indeed. It seems unlikely to me that this is pg_dump's fault; I think it's just reporting the way things are in the source database. If projects_historydata_id_seq is the wrong width, that's probably because it was manually created that way and then manually attached to projects_historydata.id.
> I suggest to run this: > ALTER SEQUENCE projects_historydata_id_seq OWNED BY public.projects_historydata.id > and then things will work better.
This won't in itself fix the sequence's width; you'd need to do some additional ALTERs if that's wrong.