Re: Checkpoint cost, looks like it is WAL/CRC

Tom,

> There's something awfully weird going on here.  I was prepared to see
> no statistically-significant differences, but not multiple cases that
> seem to be going the "wrong direction".

There's a lot of variance in the tests.   I'm currently running a variance 
test battery on one machine to figure out why the results become so 
variable when checkpointing is < 1 hour.  

Actually, to cover all which is currently running:

machine1 & 3 wal_buffers scaling test
machine2 STP variance test
machine4 full_page_writes=false test

All of these should wind up in about 5 days, provided the STP doesn't crash 
(not an insignificant concern).

-- 
--Josh

Josh Berkus
Aglio Database Solutions
San Francisco