Update "don't truncate with failsafe" rationale.
authorPeter Geoghegan <pg@bowt.ie>
Tue, 15 Feb 2022 23:16:19 +0000 (15:16 -0800)
committerPeter Geoghegan <pg@bowt.ie>
Tue, 15 Feb 2022 23:16:19 +0000 (15:16 -0800)
commit988ffc30637ce2b7c4ff984513f4695372602ce2
treefcf88f578b64ae176521e746162d79aec5b99386
parent3b0ee7f583dc347c4aecb57d35830d2f99956e5c
Update "don't truncate with failsafe" rationale.

There is a very good (though non-obvious) reason to avoid relation
truncation during a VACUUM that has triggered the failsafe mechanism,
which was missed before now.  Update related comments, so this isn't
forgotten.

Reported-By: John Naylor <john.naylor@enterprisedb.com>
Discussion: https://postgr.es/m/CAFBsxsFiMPxQ-dHZ8tOgktn=+ffeJT3+GinZ4zdOGbmAnCYadA@mail.gmail.com
src/backend/access/heap/vacuumlazy.c