On Fri, May 07, 2021 at 04:30:00PM -0400, Tom Lane wrote:
> I can certainly see an argument for running some buildfarm animals
> with fsync on (for all tests). I don't see a reason for forcing
> them all to run some tests that way; and if I were going to do that,
> I doubt that 008_fsm_truncation.pl would be the one I would pick.
> I think it's nothing but sloppiness that that one is out of step with
> all the rest.
My take on this point is that using the configuration that can be
enforced for each animal would be enough. I manage a small animal and
this stuff can take a while to flush some data.
Worth noting that using fsync=on has not been discussed on the
original thread, and I don't see why that's necessary:
https://www.postgresql.org/message-id/flat/CABOikdNr5vKucqyZH9s1Mh0XebLs_jRhKv6eJfNnD2wxTn%3D_9A%40mail.gmail.com
So I would vote for removing it in this case.
--
Michael
