On Thu, 17 Jan 2019 at 05:59, Laurenz Albe <laurenz.a...@cybertec.at> wrote:
> Tsunakawa, Takayuki wrote: > > From: Tom Lane [mailto:t...@sss.pgh.pa.us] > > > The problem here of course is that whoever invented > target_session_attrs > > > was unconcerned with following that precedent, so what we have is > > > "target_session_attrs=(any | read-write)". > > > Are we prepared to add some aliases in service of unifying these names? > > > > I think "yes". > > > > > 2. Whether or not you want to follow pgJDBC's naming, it seems like we > > > ought to have both "require read only" and "prefer read only" behaviors > > > in this patch, and maybe likewise "require read write" versus "prefer > > > read write". > I just had a look at the JDBC code there is no prefer read write. There is a "preferSecondary" The logic behind this is that the connection would presumably be only doing reads so ideally it would like a secondary, but if it can't find one it will connect to a primary. To be clear there are 4 target server types in pgJDBC, "any", "master","secondary", and "preferSecondary" (looking at this I need to alias master to primary, but that's another discussion) I have no idea where "I want to write but I'm OK if I cannot came from"? Dave
