On 12/8/15 3:52 AM, Jeremy Harris wrote:
> On 07/12/15 16:44, Simon Riggs wrote:
>> There are many optimizations we might adopt, yet planning time is a factor.
>> It seems simple enough to ignore more complex optimizations if we have
>> already achieved a threshold cost (say 10). Such a test would add nearly
>> zero time for the common case. We can apply the optimizations in some kind
>> of ordering depending upon the cost, so we are careful to balance the
>> cost/benefit of trying certain optimizations.
>
> Given parallelism, why not continue planning after initiating a
> a cancellable execution, giving a better plan to be used if the
> excecution runs for long enough?
Because that would take significantly more work than what Simon is
proposing.
That said, I think the ability to restart with a different plan is
something we might need, for cases when we discover the plan estimates
were way off. If that ever gets built it might be useful for what you
propose as well.
--
Jim Nasby, Data Architect, Blue Treble Consulting, Austin TX
Experts in Analytics, Data Architecture and PostgreSQL
Data in Trouble? Get it in Treble! http://BlueTreble.com