Re: Optimize commit performance with a large number of 'on commit delete rows' temp tables
От | wenhui qiu |
---|---|
Тема | Re: Optimize commit performance with a large number of 'on commit delete rows' temp tables |
Дата | |
Msg-id | CAGjGUAKp6H4E6GnfpVGsfD3=9fm++2nAHaKMnb-6NDTFOWFsCQ@mail.gmail.com обсуждение исходный текст |
Ответ на | Re: Optimize commit performance with a large number of 'on commit delete rows' temp tables ("feichanghong" <feichanghong@qq.com>) |
Ответы |
Re: Optimize commit performance with a large number of 'on commit delete rows' temp tables
|
Список | pgsql-hackers |
Hi feichanghong
Thanks for updating the patch ,I think could be configured as a GUC parameter,PostgreSQL has too many static variables that are written to death and explicitly stated in the code comments may later be designed as parameters. Now that more and more applications that previously used oracle are migrating to postgresql, there will be more and more scenarios where temporary tables are heavily used.Because oracle will global temporary tablespace optimised for this business scenario, which works well in oracle, migrating to pg faces very tricky performance issues,I'm sure the patch has vaule
Best Regards
Thanks for updating the patch ,I think could be configured as a GUC parameter,PostgreSQL has too many static variables that are written to death and explicitly stated in the code comments may later be designed as parameters. Now that more and more applications that previously used oracle are migrating to postgresql, there will be more and more scenarios where temporary tables are heavily used.Because oracle will global temporary tablespace optimised for this business scenario, which works well in oracle, migrating to pg faces very tricky performance issues,I'm sure the patch has vaule
Best Regards
feichanghong <feichanghong@qq.com> 于2024年7月6日周六 03:40写道:
The patch in the attachment, compared to the previous one, adds a threshold forusing the bloom filter. The current ON_COMMITS_FILTER_THRESHOLD is set to 64,which may not be the optimal value. Perhaps this threshold could be configuredas a GUC parameter?Best Regards,Fei Changhong
В списке pgsql-hackers по дате отправления: