Re: bgwriter, regression tests, and default shared_buffers settings

On 07/18/2012 03:59 PM, Tom Lane wrote:
> 2. It's rather disturbing that a fairly large swath of functionality
> just stopped getting tested at all by the buildfarm.  Do we want to
> rethink the shared_buffers increase?  Or artificially bloat the
> regression database to make it larger than 128MB?  Or do something else
> to ensure we still exercise the DB-bigger-than-buffers case?

A couple of other ideas:

The buildfarm does have the ability to set config data after initdb has 
run (which I just enhanced in the latest release). So a buildfarm owner 
could add a config line for shared_buffers which would override what 
initdb had set.

Or we could provide an initdb flag which would set an upper bound on 
shared_buffers, and have make check (at least) use it.

I'd rather not bloat the regression database if we can reasonably avoid 
it. Buildfarm members are often tight on space.

cheers

andrew