On Sat, Apr 9, 2016 at 6:57 AM, Stas Kelvich <s.kelvich@postgrespro.ru> wrote:
> Aren’t "Twophase transactions on slave” falling into category of patches that fixes
> previously introduces behaviour? |'m not trying to argue with RMT decision, but just
> want to ensure that it was thoughtful decision, taking into account that absence of that
> patch in release can cause problems with replication in some cases as it was warned
> by Jesper[1] and Andres[2].
The RMT hasn't made a specific decision on this patch. I merely moved
it in accordance with the general RMT decision about feature freeze:
http://www.postgresql.org/message-id/CA+TgmoY56w5FOzeEo+i48qehL+BsVTwy-Q1M0xjUhUCwgGW7-Q@mail.gmail.com
My personal view is as follows:
If the patch sped up things on the master but not on the slave, that
doesn't justify a post-freeze change to speed up the slave. That can
be done for 9.7. On the other hand, if the patch broke things that
are supposed to work, then that must be fixed.
--
Robert Haas
EnterpriseDB: http://www.enterprisedb.com
The Enterprise PostgreSQL Company
