Re: BUG #16732: pg_dump creates broken backups

On Fri, Nov 20, 2020 at 6:03 PM Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> wrote:

> Zsolt Ero <zsolt(dot)ero(at)gmail(dot)com> writes:
> > It happens with 1 row in a 3 GB gzip compressed database dump. I'm
> > thinking about how could I possibly give you a reproducible case. Do you
> > know any way which doesn't require me to share the whole production
> > database? (which is not an option)
>
> Of course not. Can you make it happen with a few rows of dummy data
> within the same schema?
>
>
Before doing that, have you positively confirmed that map_id=112664 exists
on the maps table in the live database? Your follow-on post is difficult
to follow. The claim about "1 record" in a database would suggest
corruption, not a structural problem - which should affect entire tables
(though you'd only see the first error). In the dump file, is 112664 the
first ID in the table data?

David J.