Re: WAL Optimisation - configuration and usage
От | Simon Riggs |
---|---|
Тема | Re: WAL Optimisation - configuration and usage |
Дата | |
Msg-id | 004901c400a3$fb570570$5baa87d9@LaptopDellXP обсуждение исходный текст |
Ответ на | Re: WAL Optimisation - configuration and usage (Josh Berkus <josh@agliodbs.com>) |
Ответы |
Re: WAL Optimisation - configuration and usage
|
Список | pgsql-performance |
>Josh Berkus wrote > >Simon Riggs wrote > > Please set WAL_DEBUG to 1 so we can see a bit more info: thanks. > > I'm pretty sure that WAL_DEBUG requires a compile-time option. In my naiveté, I just set and use it. I discovered it in the code, then set it to take advantage. I'm surprised, but you are right, the manual does SAY this requires a compile time option; it is unfortunately not correct. Maybe this was once as you say? If Rob is using 7.2.2 maybe this still applies to him? I dont know. Setting wal_debug > 0 in postgresql.conf sets a variable XLOGDEBUG, which although it is all capitals is not in fact a compiler directive, as it looks. This variable is used within xlog.c to output **too much** information to the log. However, it is the only option at present. This prompts me however to consider the idea of having various levels of WAL debug output, or using some kind of log_checkpoint mechanism to better understand what is going on. Best Regards, Simon Riggs
В списке pgsql-performance по дате отправления: