Re: Hash partitioning.

<div dir="ltr">On Tue, Jun 25, 2013 at 12:08 PM, Yuri Levinsky <span dir="ltr"><<a href="mailto:yuril@celltick.com"
target="_blank">yuril@celltick.com</a>></span>wrote:<br /><div class="gmail_extra"><div
class="gmail_quote"><blockquoteclass="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc
solid;padding-left:1ex">Guys,<br/> I am sorry for taking your time. The reason for my question is:<br /> As former
OracleDBA and now simple beginner PostgreSQL DBA I would like<br /> to say: the current partitioning mechanism might be
improved.Sorry, it<br /> seems to me far behind yesterday requirements. As model for improvement<br /> the Oracle might
betaken as example. Unfortunately I am not writing an<br /> C code and see my benefit to PostgreSQL community in only
risingthis<br /> issue. I'll be very happy to be helpful in something else, but...<br /></blockquote></div><br
/></div><divclass="gmail_extra">Please don't flee over this...<br /><br /></div><div class="gmail_extra">As I think you
cansee, now, the partitioning problem is tougher than it<br />may at first seem to be.  It's quite useful to quickly
getto the point of<br /> understanding that.<br /><br /></div><div class="gmail_extra">There would indeed be merit in
improvingthe partitioning apparatus,<br /></div><div class="gmail_extra">and actually, I think it's been a couple of
yearssince there has been<br /> serious discussion of this.<br /><br /></div><div class="gmail_extra">The discussion
tendsto head into the rabbit hole of disputing about<br />whether one mechanism or another is ideal.  That's the wrong
starting<br/>point - we shouldn't start with what's easiest to make "ideal," we <br /> should start by determining what
isrequired/desirable, without too<br />much reference, at least initially, on to how to achieve it.<br /></div><div
class="gmail_extra">--<br />When confronted by a difficult problem, solve it by reducing it to the<br /> question, "How
wouldthe Lone Ranger handle this?"<br /></div></div>