When I execute the query below on a Sybase Client, 518 rows are returned.
When I execute the query on QlikView more than 15 millions rows are returned.
It seems like QlikView is trying to recover the entire tables before joining
Any clue?
SELECT DISTINCT CONVERT(DATE, a.inicio_atendente_status) AS data_inicio, c.id_operacao, b.id_usuario, a.id_status, CONVERT(TIME, a.inicio_atendente_status) AS inicio_status, CONVERT(TIME, a.fim_atendente_status) AS fim_status, DATEDIFF(ms, a.inicio_atendente_status, a.fim_atendente_status) AS duracao_status, d.nome_status, a.inicio_atendente_status, a.fim_atendente_status, CASE WHEN DATEPART(dd,a.inicio_atendente_status) = DATEPART(dd,a.fim_atendente_status) THEN 'OK' ELSE 'FALSE' END FROM tb_atendente_status_historico a, tb_equipe_usuario b, tb_equipe c, tb_status d WHERE a.id_equipe_usuario = b.id_equipe_usuario AND a.id_status <> 40 AND a.inicio_atendente_status BETWEEN '2010-05-12 00:00:00.000' AND '2010-05-12 23:59:59.999' AND c.id_equipe = b.id_equipe AND c.id_operacao in (17) AND a.id_status = d.id_status ORDER BY CONVERT(date, a.inicio_atendente_status), b.id_usuario, a.id_status
