On Sat, Jun 13, 2020 at 09:27:25PM -0400, Bruce Momjian wrote:
> On Sat, Jun 13, 2020 at 05:06:37PM -0500, Ron wrote:
> > On 6/13/20 1:46 PM, Bruce Momjian wrote:
> > > On Wed, Jun 3, 2020 at 08:53:45PM +0200, Andreas Joseph Krogh wrote:
> > > > I agree these are all technical issues, but nevertheless - "implementation
> > > > details", which DBAs don't care about. What's important from a DBA's
> > > > perspective is not whether WAL is cluster-wide or database-wide, but whether
> > > > it's possible to manage backups/PITR/restores of individual databases in a more
> > > > convenient matter, which other RDBMS-vendors seem to provide.
> > > > I love PG, have been using it professionally since 6.5, and our company depends
> > > > on it, but there are things other RDBMS-vendors do better...
> > > The bigger issue is that while we _could_ do this, it would add more
> > > problems and complexity, and ultimately, I think would make the
> > > software less usable overall and would be a net-negative. We know of no
> > > way to do it without a ton of negatives.
> >
> > How do other RDBMSs do it with ease? (I know it's an architectural issue,
> > but what's the architectural issue?)
>
> I don't know.
I don't know the details, but I do know the general issues. Other
vendors must have sacrificed architectural simplicity, features,
reliability, or performance to allow these things. For example, it
wouldn't be hard to just make databases another level of container above
schemas to allow for simple cross-database queries, but we would lose
the security isolation of databases (connection control. private system
tables and extensions) to do that. Having per-database WAL causes loss
of performance, reliability issues, and architectural complexity. Those
problems might be solvable, but you will need to take a hit in one of
these areas.
--
Bruce Momjian <bruce@momjian.us> https://momjian.us
EnterpriseDB https://enterprisedb.com
The usefulness of a cup is in its emptiness, Bruce Lee