On Thu, 22 Sept 2022 at 15:16, Alvaro Herrera <alvhe...@alvh.no-ip.org> wrote: > > On 2022-Sep-22, Simon Riggs wrote: > > > On Mon, 19 Sept 2022 at 00:16, Greg Stark <st...@mit.edu> wrote: > > > > VACUUM was willing to remove a committed-dead tuple immediately if it > > > was > > > deleted by the same transaction that inserted it. The idea is that > > > such a > > > tuple could never have been visible to any other transaction, so we > > > don't > > > need to keep it around to satisfy MVCC snapshots. However, there was > > > already an exception for tuples that are part of an update chain, and > > > this > > > exception created a problem: we might remove TOAST tuples (which are > > > never > > > part of an update chain) while their parent tuple stayed around (if > > > it was > > > part of an update chain). This didn't pose a problem for most things, > > > since the parent tuple is indeed dead: no snapshot will ever consider > > > it > > > visible. But MVCC-safe CLUSTER had a problem, since it will try to > > > copy > > > RECENTLY_DEAD tuples to the new table. It then has to copy their > > > TOAST > > > data too, and would fail if VACUUM had already removed the toast > > > tuples. > > > Good research Greg, thank you. Only took 10 years for me to notice it > > was gone ;-) > > But this begs the question: is the proposed change safe, given that > ancient consideration? I don't think TOAST issues have been mentioned > in this thread so far, so I wonder if there is a test case that verifies > that this problem doesn't occur for some reason.
Oh, completely agreed. I will submit a modified patch that adds a test case and just a comment to explain why we can't remove such rows. -- Simon Riggs http://www.EnterpriseDB.com/