Re: BUG #15636: PostgreSQL 11.1 pg_basebackup backup to a CIFSdestination throws fsync error at end of backup

On Mon, Feb 18, 2019 at 3:55 PM Michael Paquier <michael@paquier.xyz> wrote:
> On Mon, Feb 18, 2019 at 02:34:15PM +1300, Thomas Munro wrote:
> > ...  Considering that, the lack of other complaints, and the
> > availability of a workaround (--no-sync followed by "sync"), I'm not
> > so sure it's worth committing this.
>
> If we get to ignore EINVAL, then we won't know that fsync has failed
> even in cases where the caller has done an incorrect thing by using a
> special file, which sounds like a bad idea for the core code as much
> as any plugins calling that?

Well the theory is that it's OK because it's only for directories (the
same case in which case we already tolerate EBADF (for an OS that
nobody documented so we don't really know if/when we can take it
out)).  But I agree with you that the patch does not spark joy.

-- 
Thomas Munro
http://www.enterprisedb.com