Re: [WIP] cache estimates, cache access cost

On Tue, Jun 14, 2011 at 6:17 PM, Greg Smith <greg(at)2ndquadrant(dot)com> wrote:
> Who said anything about this being a commit candidate?  The "WIP" in the
> subject says it's not intended to be.  The community asks people to submit
> design ideas early so that ideas around them can be explored publicly.  One
> of the things that needs to be explored, and that could use some community
> feedback, is exactly how this should be benchmarked in the first place.
>  This topic--planning based on cached percentage--keeps coming up, but
> hasn't gone very far as an abstract discussion.  Having a patch to test lets
> it turn to a concrete one.
>
> Note that I already listed myself as the reviewer  here, so it's not even
> like this is asking explicitly for a community volunteer to help.  Would you
> like us to research this privately and then dump a giant patch that is
> commit candidate quality on everyone six months from now, without anyone
> else getting input to the process, or would you like the work to happen
> here?  I recommended Cédric not ever bother soliciting ideas early, because
> I didn't want to get into this sort of debate.  I avoid sending anything
> here unless I already have a strong idea about the solution, because it's
> hard to keep criticism at bay even with that.  He was more optimistic about
> working within the community contribution guidelines and decided to send
> this over early instead.  If you feel this is too rough to even discuss,
> I'll mark it returned with feedback and we'll go develop this ourselves.

My usual trope on this subject is that WIP patches tend to elicit
helpful feedback if and only if the patch author is clear about what
sort of feedback they are seeking. I'm interested in this topic, so,
I'm willing to put some effort into it; but, as I've said before, I
think this patch is coming from the wrong end, so in the absence of
any specific guidance on what sort of input would be useful, that's
the feedback you're getting. Feel free to clarify what would be more
helpful. :-)

Incidentally, I have done a bit of math around how to rejigger the
costing formulas to take cached_page_cost and caching_percentage into
account, which I think is the most interesting end place to start this
work. If it's helpful, I can write it up in a more organized way and
post that; it likely wouldn't be that much work to incorporate it into
what Cedric has here already.

--
Robert Haas
EnterpriseDB: http://www.enterprisedb.com
The Enterprise PostgreSQL Company