On Thu, Sep 12, 2019 at 12:25 PM Rodrigo Ramírez Norambuena
<decipher.hk@gmail.com> wrote:
>
> On Wed, Sep 11, 2019 at 4:05 PM Jonathan S. Katz <jkatz@postgresql.org> wrote:
> >
>
>
> > For 0005, I would not put that in a migration. I would put that in a
> > test runner or whatever script we will use to build the test
> > environment, to mock varnish (if we need to -- not sure if we do?).
>
> Mock the varnish in first step could be more obfuscate because a
> purge_urls is present in the almost all models. The kick start is a
> add sql into db as varnish_local, after that could add a helper to
> test varnish function to load these functions in db.
>
> About not put in migration seams good choice but this migration run
> only in test model. I'll could find a other way to add for the runner
> or more clean. If you have any idea about this let me know.
For this topic, I attach a new patch with a runner. Inside the runner
is execute SQL sentences for Varnish.
This is better approach against to previous inside of migration model.
Regards!
--
Rodrigo Ramírez Norambuena
http://www.rodrigoramirez.com/