Re: Driver behaves differently with prepareThreshold and timestamp fields when daylights is active (was Re: Re: 9.4-1207 behaves differently with server side prepared statements compared to 9.2-1102)
От | Edson Richter |
---|---|
Тема | Re: Driver behaves differently with prepareThreshold and timestamp fields when daylights is active (was Re: Re: 9.4-1207 behaves differently with server side prepared statements compared to 9.2-1102) |
Дата | |
Msg-id | BLU437-SMTP365486009F27D05E49B558CFC90@phx.gbl обсуждение исходный текст |
Ответ на | Driver behaves differently with prepareThreshold and timestamp fields when daylights is active (was Re: Re: 9.4-1207 behaves differently with server side prepared statements compared to 9.2-1102) (Edson Richter <edsonrichter@hotmail.com>) |
Ответы |
Re: Driver behaves differently with prepareThreshold and timestamp
fields when daylights is active (was Re: Re: 9.4-1207 behaves
differently with server side prepared statements compared to 9.2-1102)
|
Список | pgsql-jdbc |
Warning: This only happens if your country has Daylights Saving Time active. In Brazil, default timezone is GMT-03:00. Currently in DST, the timezone is GMT-02:00. This seems to affect some kind of value conversion, and subtract 1 hour from timestamp values. Atenciosamente, Edson Carlos Ericksson Richter Em 11/01/2016 14:15, Edson Richter escreveu: > I'm using 1201 driver. No matter JDK 7 or JDK 8. > Big real issue for me: after third query, the prepareThresold is hit, > and timezone data is converted "on the fly" to "timezone without > daylights" for timestamp fields. > Data is stored in database as "2015-09-30 00:00:00", in the 1st to 3rd > query returns with "2015-09-30 00:00:00", and the 4º and on returns > "2015-09-29 23:00:00". > > Disabling with prepareThreshold=0 solved the problem once for all. > I had no further time to investigate and/or produce a case for asking > for a fix. > > Regards, > > Atenciosamente, > > Edson Carlos Ericksson Richter > > Em 11/01/2016 12:14, Vladimir Sitnikov escreveu: >>> In my case, problema is that after optimization at server side, >>> results get different. >> Do you mean "wrong results" kind of issue? >> which driver version are you using? >> >> Generally speaking, it is worth submitting "bad performance when using >> prepared statements" issues to PostgreSQL hackers team. >> If just a couple of SQLs behave badly due to server-prepared >> statements, then it might make sense just tune the statements in >> question. >> Vladimir >> >> > > >
В списке pgsql-jdbc по дате отправления: