Re: What limits Postgres performance when the whole database lives in cache?

On Fri, Sep 2, 2016 at 8:36 PM, dandl <david(at)andl(dot)org> wrote:
> The paper is substantially in agreement with the presentation I quoted. If there are differences in detail, they certainly don't dominate his argument.

My point is that the paper is rather light on details of the kind that
are really important. And, that it's noteworthy that Stonebraker has
in the past, during presentations, emphasized the buffer lock
crabbing/latch coupling thing *at length*, even though it's a totally
solved problem.

It's also true that Postgres has become vastly more scalable in the
past few years due to optimization that doesn't change the fundamental
nature of the system at all, so it's very easy to imagine individual
differences being more important than differences between major
classes of system.

Those are facts. You may take from them what you will.

> IMO your claim is far weaker. What specifically do you say is wrong about his current claims, and on what facts to you base it?

I'm not the one making overarching conclusions. I'm not trying to
convince you of anything.

--
Peter Geoghegan