čt 22. 8. 2019 v 3:11 odesílatel Jeff Janes napsal:
> On Tue, Aug 20, 2019 at 11:12 AM Felix Geisendörfer
> wrote:
> ...
>
>
>> [1] My actual query had bad estimates for other reasons (GIN Index), but
>> that's another story. The query above was of course deliberately designed
>> to have bad es
On Tue, Aug 20, 2019 at 11:12 AM Felix Geisendörfer
wrote:
...
> [1] My actual query had bad estimates for other reasons (GIN Index), but
> that's another story. The query above was of course deliberately designed
> to have bad estimates.
>
As noted elsewhere, v12 thwarts your attempts to deli
Em 21/08/2019 04:30, Barbu Paul - Gheorghe escreveu:
I wonder how I missed that... probabily because of the "WHERE" clause
in what I already had.
I indexed by scheduler_task_executions.device_id and the new plan is
as follows: https://explain.depesz.com/s/cQRq
Can it be further improved?
Limit
I wonder how I missed that... probabily because of the "WHERE" clause
in what I already had.
I indexed by scheduler_task_executions.device_id and the new plan is
as follows: https://explain.depesz.com/s/cQRq
Can it be further improved?
Limit (cost=138511.45..138519.36 rows=2 width=54) (actual
t
