Re: Patch: Implement failover on libpq connect level.

On 6 November 2015 at 13:34, Robert Haas <robertmhaas@gmail.com> wrote:

>> But some options control how
>> next host should be choosen (i.e. use random order for load-balancing
>> or sequential order for high availability), so they should be specified
>> only once per connect string.
>
> But this seems like a point worthy of consideration.

This makes me think that trying to wedge this into the current API
using a funky connection string format might be a mistake.

Lots of these issues would go away if you could provide more than just
a connstring.

>> Third category of options are specified per-cluster much more often
>> than per node. But are quite often changed from compiled in default.
>
> This, too.

Like this. If you have a global set of connection options, then
per-connection options, it's a lot simpler.

I guess that can be hacked in with a more dramatic change in the
connstring format, otherwise, incorporating subsections or something.
I'm not keen on doing anything like that, but there are all sorts of
options...

"global:[user=fred port=5432] host1:[host=somehost user=user1]
host2:[host=localhost]"


(puts head in brown paper bag)

-- 
 Craig Ringer                   http://www.2ndQuadrant.com/
 PostgreSQL Development, 24x7 Support, Training & Services