Horton

Re: [HACKERS] Slow count(*) again...

Scott Marlowe <scott.marlowe@gmail.com>

View thread
From: Scott Marlowe <scott.marlowe@gmail.com>
To: Robert Haas <robertmhaas@gmail.com>
Cc: david@lang.hm, Vitalii Tymchyshyn <tivv00@gmail.com>, Kenneth Marshall <ktm@rice.edu>, Jon Nelson <jnelson+pgsql@jamponi.net>, Andrew Dunstan <andrew@dunslane.net>, Bruce Momjian <bruce@momjian.us>, Tom Lane <tgl@sss.pgh.pa.us>, Craig Ringer <craig@postnewspapers.com.au>, "pgsql-performance@postgresql.org" <pgsql-performance@postgresql.org>
Date: 2011-02-05T08:38:40Z
Lists: pgsql-hackers, pgsql-performance 
On Fri, Feb 4, 2011 at 11:37 PM, Robert Haas <robertmhaas@gmail.com> wrote:
> On Sat, Feb 5, 2011 at 12:46 AM,  <david@lang.hm> wrote:
>>> Actually for me the main "con" with streaming analyze is that it adds
>>> significant CPU burden to already not too fast load process. Especially if
>>> it's automatically done for any load operation performed (and I can't see
>>> how it can be enabled on some threshold).
>>
>> two thoughts
>>
>> 1. if it's a large enough load, itsn't it I/O bound?
>
> Sometimes.  Our COPY is not as cheap as we'd like it to be.

With a 24 drive RAID-10 array that can read at ~1GB/s I am almost
always CPU bound during copies.  This isn't wholly bad as it leaves
spare IO for the rest of the machine so regular work carries on just
fine.