Re: Syscaches should store negative entries, too

Hannu Krosing <hannu@krosing.net> writes:
> Are there _any_ tests where it does start to drop stuff ?

> In other words - is the stuff-dropping part tested reasonably recently
> (or at all) ?

Good point.  The drop mechanism is exercised well enough as driven by
shared-cache-invalidation cases, but the case where it's driven by cache
entry count isn't getting tested.

> In other words we should cache Frequently Asked Questions and not
> Frequently Found Answers ;)

Right ;-)
        regards, tom lane