On 5/23/13 7:56 PM, Claudio Freire wrote:
> Besides of the obvious option of making a lighter check (doesn't have
> to be 100% precise), wouldn't this check be done when it would
> otherwise sleep? Is it so heavy still in that context?
A commit to typical 7200RPM disk is about 10ms, while
autovacuum_vacuum_cost_delay is 20ms. If the statement cost limit logic
were no more complicated than commit_delay, it would be feasible to do
something similar each time a statement was being put to sleep.
I suspect that the cheapest useful thing will be more expensive than
commit_delay's test. That's a guess though. I'll have to think about
this more when I circle back toward usability. Thanks for the
implementation idea.
--
Greg Smith 2ndQuadrant US greg@2ndQuadrant.com Baltimore, MD
PostgreSQL Training, Services, and 24x7 Support www.2ndQuadrant.com