On Mon, Feb 20, 2017 at 11:52 AM, Michael Paquier
<michael.paquier@gmail.com> wrote:
> On Thu, Feb 16, 2017 at 10:55 AM, Michael Paquier
> <michael.paquier@gmail.com> wrote:
>> It is possible to get a test easily in this area by abusing of the
>> fact that multiple -d switches defined in psql make it use only the
>> last value. By looking at psql() in PostgresNode.pm you would see what
>> I mean as -d is defined by default. Or we could just enforce
>> PGHOST/PGPORT as there is a method to get the unix domain directory.
>> Not sure which one of those two methods is most elegant though. I
>> would tend for just using PGHOST and PGPORT.
>
> What do you think about the patch attached then? As env{PGHOST} is set
> once and for all in INIT for each test run, I have gone with the
> approach of building connection strings and feed that to psql -d. I
> have reused 001_stream_rep.pl so as connections are done on already
> existing nodes, making the test really cheap. Here is how the tests
> look:
> # testing connection parameter target_session_attrs
> ok 5 - connect to node master if mode "read-write" and master,standby_1 listed
> ok 6 - connect to node master if mode "read-write" and standby_1,master listed
> ok 7 - connect to node master if mode "any" and master,standby_1 listed
> ok 8 - connect to node standby_1 if mode "any" and standby_1,master listed
>
> I have registered an entry in the CF here:
> https://commitfest.postgresql.org/13/1014/
Thanks! Committed.
--
Robert Haas
EnterpriseDB: http://www.enterprisedb.com
The Enterprise PostgreSQL Company