Simon Riggs wrote:
> Based upon that series of conversations, I've reworked the design so
> that there is (currently) only a single standby offering sync rep at any
> one time. Other standbys can request to be sync standbys but they only
> become the sync standby if the first one fails. Which was simple to do
> and bridges the challenges of an exactly identified sync standby and the
> fragility of too closely specifying the config.
>
That seems like a good enough starting point to cover a lot of cases.
Presuming the two servers each at two sites config that shows up in a
lot of these discussions, people in the "I need sync to a remote spot"
can get that, and if that site is unavailable for long enough to be
kicked out they'll smoothly degrade to sync on a secondary local copy.
And those who want high performance local sync and best-effort for the
remote site can configure for that too, and if the local secondary dies
they'll just degrade to the slow remote commits.
--
Greg Smith 2ndQuadrant US greg@2ndQuadrant.com Baltimore, MD
PostgreSQL Training, Services and Support www.2ndQuadrant.us
"PostgreSQL 9.0 High Performance": http://www.2ndQuadrant.com/books
