On Tue, Oct 17, 2017 at 2:29 PM, Glenn Pierce <glennpie...@gmail.com> wrote:
> and I have a simple query that fails > This is not failure, this is a query that found zero matching records. > > Ie > > SELECT sensor_id, MAX(ts), date_trunc('day', ts), COALESCE(MAX(value), > 'NaN')::float FROM sensor_values_days WHERE ts > '2017-10-06 > 00:01:01+00' AND ts < '2017-10-06 23:59:59+00' GROUP BY 1, 3 ORDER BY > 1, 2; > sensor_id | max | date_trunc | coalesce > -----------+-----+------------+---------- > (0 rows) > > > If I remove the timezone part of the start date I get results. > > Ie > > > ts > '2017-10-06 00:01:01' > > 597551 | 2017-10-06 01:00:00+01 | 2017-10-06 00:00:00+01 | 13763 > > I'm sure I am doing something silly but can't see what. Does anyone know what is going on here ? > The "max(ts)" result indicates a time of midnight, the 6th, GMT ts > '2017-10-06 00:01:01+01' equates to > '2017-10-05 23:01:01+00' of which midnight, the 6th, GMT is indeed more recent ts > '2017-10-06 00:01:01+00' is 12:01:01 on the 6th, GMT, of which midnight GMC, the 6th is NOT more recent David J.