On Thu, Jan 24, 2019 at 6:42 PM Thomas Munro
<thomas.munro@enterprisedb.com> wrote:
> Yes, this is essentially the same thing that you were arguing against
> above. Perhaps you are right, and there are no people who would want
> synchronous replay, but not synchronous commit.
Maybe I'm misunderstanding the terminology here, but if not, I find
this theory wildly implausible. *Most* people want read-your-writes
behavior. *Few* people want to wait for a dead standby. The only
application of the later is when even a tiny risk of transaction loss
is unacceptable, but the former has all kinds of clustering-related
uses.
--
Robert Haas
EnterpriseDB: http://www.enterprisedb.com
The Enterprise PostgreSQL Company