Re: Bugs/slowness inserting and indexing cubes

Robert Haas wrote:
> On Mon, Feb 13, 2012 at 7:45 AM, Robert Haas<robertmhaas(at)gmail(dot)com> wrote:
>> On Thu, Feb 9, 2012 at 3:37 PM, Jay Levitt<jay(dot)levitt(at)gmail(dot)com> wrote:
>>> So my pre-built 9.1.2 takes 434s, my source-built 9.2 takes 509s, and
>>> (probably both of our) 9.1-HEAD takes 1918s... is that something to worry
>>> about, and if so, are there any tests I can run to assist? That bug doesn't
>>> affect me personally, but y'know, community and all that. Also, I wonder if
>>> it's something like "9.2 got way faster doing X, but meanwhile, HEAD got way
>>> slower doing Y.", and this is a canary in the coal mine.
>> This might be a lame hypothesis, but... is it possible that you built
>> your 9.1-tip binaries with --enable-cassert? Or with different
>> optimization options?

No, I think I/O just varies more than my repeated tests on 1M rows
indicated. I ran the 10M-row test four times on the same server,
alternating between packaged 9.1.2 and source-built 9.1.2 (default configure
options), and saw these times:

INSERT INDEX
apt 76 578
source 163 636
apt 73 546
source 80 473

EBS has no performance guarantees at all; you share your disks with an
arbitrary number of other users, so if someone "in the neighborhood" decides
to do some heavy disk I/O, you lose. Let this be a lesson to me: run
benchmarks locally!

> So I tested. On my MacBook Pro, your test script builds the index in
> just over 25 s on both REL9_1_2 and this morning's REL9_1_STABLE.

I think that's the 1-million version I emailed; try adding a zero and see if
it doesn't take a little longer.

Jay