Jim Nasby <Jim.Nasby@BlueTreble.com> writes:
> On 10/31/15 11:19 AM, Nathan Wagner wrote:
>> I think it would be very useful to just be able to tell the system "fire
>> this up for me so I can test it". I don't think it needs to handle
>> every possible testing scenario, just making it easier to leave up the
>> test postmaster from make check would be very useful, at least to me.
> I've wished that the cluster setup and teardown behavior of pg_regress
> was available outside of pg_regress itself. Would that mostly suffice
> for what you're looking for?
I should think not. pg_regress is not merely not encouraging of outside
connections to the started postmaster; if such is even possible, it's
likely to be regarded as a security bug. What Nathan is looking for is
arguably useful, but I do not think pg_regress should be expected to
support it. It needs to be a different tool, with different security
parameters.
regards, tom lane