Andrew Dunstan <andrew@dunslane.net> writes:
> Here is some more info. Below is a trace from dropdb. There is a loop
> around the rmdir() calls which I have set to time out at 600 seconds.
> The call eventually succeeds after around 300 seconds (I've seen this
> several times). It looks like we are the victim of some caching - the
> directory still thinks it has some of the files it has told us we have
> deleted successfully.
If you rescan the directory after deleting the files, does it show
as empty?
> Bottom line, this is a real mess. Surely postgres is not the only
> application in the world that wants to be able to delete a directory
> tree reliably on Windows. What do other apps do?
I'm wondering if this is a side effect of the way win32_open does
things. It's hard to believe that rmdir is that bogus in general,
but perhaps win32_open is causing us to exercise a corner case?
regards, tom lane