Re: PsqlODBC slow on UNION queries
От | Zoltan Boszormenyi |
---|---|
Тема | Re: PsqlODBC slow on UNION queries |
Дата | |
Msg-id | 43C4D590.5090704@dunaweb.hu обсуждение исходный текст |
Ответ на | Re: PsqlODBC slow on UNION queries (Ludek Finstrle <luf@pzkagis.cz>) |
Ответы |
Re: PsqlODBC slow on UNION queries
|
Список | pgsql-odbc |
Ludek Finstrle írta: >>>I forgot. Feel free to send the mylog output here to list. We can >>>help you to decode it ;-) >>> >>> >>I will, when I am back to my workplace. >> >>In the meantime, I have set MyLog on in WinXP under VMWare here, >>but Access produced some 120+ MB log and it wasn't even finished >>reading the records... I didn't feel like I have to send it to the list. ;-) >>Anyway, it turned out it used DECLARE CURSOR and it fetched >>100 records at a time. >> >> > >Ufff. What tons of records do you get? BTW could you compile psqlODBC > > About 86000 records and I found every fields in every records logged, too. >from source code? It is possible to disable mylog output in function > > I can, but not on Windows. I can take the postgresql-odbc src.rpm from RedHat Rawhide and replace the source TGZ with the current CVS snapshot. In fact, I have done it several times, to test my Linux apps against the "official" ODBC driver instead of the ones that come with unixODBC. But that doesn't help me under Windows. :-) I don't have VC++. >CC_mapping (it's per record mylog for problems with encoding). I think >you don't have such a problem so there is no problem to comment out >mylog(...) calls from CC_mapping. This results in much smaller mylog >output. > >Regards, > >Luf > > > Best regards, Zoltán Böszörményi
В списке pgsql-odbc по дате отправления: