On Tue, Mar 26, 2019 at 5:10 PM Tomas Vondra <tomas.von...@2ndquadrant.com> wrote: > Bogus might be a bit too harsh, but yeah - failure to reliably detect > obviously > invalid checksums when the LSN just happens to be high due to randomness is > not > a good thing. We'll still detect pages corrupted in other places, but this is > rather unfortunate.
I have personally seen real world corruption that involved a page image consisting of random noise. Several times. Failing to detect blatant corruption is unacceptable IMV. Can't we do better here without great difficulty? There are plenty of generic things that you we could do that can verify that almost any type of initialized page is at least somewhat sane. For example, you can verify that line pointers indicate that tuples are non-overlapping. That said, Andres' approach sounds like the way to go to me. -- Peter Geoghegan
