On Wed, 2008-01-09 at 16:20 +0100, Markus Schiltknecht wrote:
> Simon Riggs wrote:
> >> I have to admit I always found it kludgy to have objects named
> >> invoices_2000_JAN and invoices_2000_FEB and so on. It's kind of an meta
> >> denormalization. But so is specifying where clauses repeatedly.
> >
> > The idea for using the WHERE clauses was to specifically avoid naming.
>
> I understand, and I'm all for avoiding needless, kludgy names.
>
> As I pointed out, knowledge of split points might be important for the
> database system. Maybe we can store the split point without the need for
> names? Dunno.
>
> > If you guys really want names, we can have names, but I think I want to
> > see a case where the storage characteristics of the table are so complex
> > we can only make sense of it by naming particular chunks.
>
> Well, assuming you only have to deal with one split point, that's
> certainly true. However, there are people using more than two table
> spaces, thus obviously needing more split points.
>
> Can we name the split points, rather than the partitions?
So far, I've been looking at partition exclusion as the most important
feature for the VLDB use case.
You seem to have moved straight on to what I've regarded as later
features for partitioning. From my side, if I can't make SE work then
most of the other features seem moot, even though I personally regard
them as important also.
With that in mind, can I clarify what you're thinking, please?
1) the things you've been discussing are so important I should do them
first, which would necessarily require named chunks of tables
2) the things you've been discussing are essential requirements of
partitioning and we could never consider it complete until they are also
included and we must therefore talk about them now to check that its all
possible before we do anything on SE
3) doing SE first is right, I'm just thinking ahead
4) the topics aren't really linked and I'm suggesting doing development
on them in parallel
or...
Sorry if that seems blunt, I'm just not clear where we're going. I have
to think about implementability, planning and priorities if I'm to get
it done.
-- Simon Riggs 2ndQuadrant http://www.2ndQuadrant.com