On Mon, Feb 25, 2019 at 08:17:27PM +0200, David Steele wrote:
> Here's the really obvious bad thing: if users do not update their procedures
> and we ignore backup_label.pending on startup then they will end up with a
> corrupt database because it will not replay from the correct checkpoint. If
> we error on the presence of backup_label.pending then we are right back to
> where we started.
Not really. If we error on backup_label.pending, we can make the
difference between a backend which has crashed in the middle of an
exclusive backup without replaying anything and a backend which is
started based on a base backup, so an operator can take some action to
see what's wrong with the server. If you issue an error, users can
also see that their custom backup script is wrong because they forgot
to rename the flag after taking a backup of the data folder(s).
The real pain point about the non-exclusive APIs is really that we
need to keep the connection opened, which can be done easily using
drivers like psycopg2 or such, still that's far from straight-forward
for the average enterprise user, and it can be hard for some companies
to complicate a software stack with more external dependencies. So
yes, while I can see the long-term benefits of non-exclusive backups,
they are much more complicated to use than exclusive backups.
--
Michael