Re: pgbench: Skipping the creating primary keys after initialization

On Wed, Aug 2, 2017 at 5:50 PM, Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> wrote:
> Robert Haas <robertmhaas(at)gmail(dot)com> writes:
>> On Wed, Aug 2, 2017 at 10:17 AM, Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> wrote:
>>> Or in other words, this looks to me quite a bit like the hackery
>>> that resulted in pgbench's -S and -N options, before we figured out
>>> that making it scriptable was a better answer.
>
>> But it's not very clear to me how we could make this case scriptable,
>
> Well, I'm imagining that "-i" would essentially become a short form
> of "-b initialize", as already happened for -S and -N, where the script
> looks something like

Yes, I would imagine a facility where one could do pgbench $script and
issue a complete test set. Here is for example a funky idea: let's
separate each script with a set of meta-commands, \init being what is
used just for initialization, and then use \script to define a set of
commands with a custom weight. Say:
\init
CREATE TABLE foo(a int);
\script select_query [weight N]
SELECT count(*) FROM foo;
\script insert_query [weight N]
INSERT INTO foo VALUES ('1');

That may be over-engineering things, but personally I don't like much
having just a switch to remove indexes. Next time we will come with
another option that only selects a portion of the indexes created.
--
Michael