Re: Clock sweep not caching enough B-Tree leaf pages?

On Wed, Apr 16, 2014 at 1:07 PM, Peter Geoghegan <pg(at)heroku(dot)com> wrote:
> On Wed, Apr 16, 2014 at 10:56 AM, Robert Haas <robertmhaas(at)gmail(dot)com> wrote:
>> I don't think he's being unreasonable, and I don't understand why
>> you're getting bent out of shape about it. You proposed a patch, he
>> articulated a problem, you don't want to fix it right now. All of
>> which is fine. Why the ad hominem accusations?
>
> I just think it's bad form to hold something like this to the same
> standards as a formal commitfest submission. I am well aware that the
> patch probably has several scalability issues.

In fairness to Andres, while *you* may know that issuing an expensive
syscall in a tight loop is on the list of Forbidden Things, a lot of
people don't and it's pretty reasonable to issue methodology
objections in order to get them documented.

Anyways, I'm still curious if you can post similar numbers basing the
throttling on gross allocation counts instead of time. Meaning: some
number of buffer allocations has to have occurred before you consider
eviction. Besides being faster I think it's a better implementation:
an intermittently loaded server will give more consistent behavior.

merlin