getGeneratedKeys method
От | Ceki Gulcu |
---|---|
Тема | getGeneratedKeys method |
Дата | |
Msg-id | 20040526115632.87009.qmail@web51008.mail.yahoo.com обсуждение исходный текст |
Ответы |
Re: getGeneratedKeys method
Re: getGeneratedKeys method |
Список | pgsql-jdbc |
Hello, We, the log4j developers, are working on an appender that will write log4j events to various RDBMS Various fields of an event object will go to 3 different tables, namely the logging_event table, the logging_event_property table and the logging_event_exception table. Each entry in the logging_event_property and the logging_event_exception tables uses a reference to event_id, where event_id is a database generated primary key in the logging_event table. For entries in the logging_event_property and logging_event_exception tables to be meaningful, we absolutely need the generated event_id each time we insert a new logging_event row. We are able to do this by following each logging_event insertion with a query to the database asking for the event_id of the last inserted event. This works for multiple database systems. However, we have discovered that batching multiple insertions gives a very significant boost in performance. Thus, we would like to insert say 50 logging_event rows, then multiple logging_event_property rows and then multiple logging_event_exception rows. We are using the JDBC getGeneratedKeys method to obtain the event_ids. Unfortunately, this function is not implemented in Postgresql. Looking at the archives it seems that this functionality requires support for the database back end which does not exist currently. Is that correct? Is there another way to insert into multiple tables in batches as described above without using the JDBC getGeneratedKeys method? Your views on the matter would be highly appreciated. Thanking you in advance, -- Ceki G�lc� For log4j documentation consider "The complete log4j manual" ISBN: 2970036908 http://www.qos.ch/shop/products/clm_t.jsp __________________________________ Do you Yahoo!? Friends. Fun. Try the all-new Yahoo! Messenger. http://messenger.yahoo.com/
В списке pgsql-jdbc по дате отправления: