Re: PATCH: regular logging of checkpoint progress
От | Magnus Hagander |
---|---|
Тема | Re: PATCH: regular logging of checkpoint progress |
Дата | |
Msg-id | CABUevEzNZmg6VHerp1=jDP5OY=XAtWGthHvvNKeEB49H8EaAkw@mail.gmail.com обсуждение исходный текст |
Ответ на | Re: PATCH: regular logging of checkpoint progress ("Tomas Vondra" <tv@fuzzy.cz>) |
Ответы |
Re: PATCH: regular logging of checkpoint progress
|
Список | pgsql-hackers |
<p><br /> On Sep 2, 2011 5:02 PM, "Tomas Vondra" <<a href="mailto:tv@fuzzy.cz">tv@fuzzy.cz</a>> wrote:<br /> ><br/> > On 2 Září 2011, 15:44, Peter Eisentraut wrote:<br /> > > On fre, 2011-09-02 at 11:01 +0200, Tomas Vondrawrote:<br /> > >> What about logging it with a lower level, e.g. NOTICE instead of the<br /> > >>current LOG? If that's not a solution then a new GUC is needed I<br /> > >> guess.<br /> > ><br />> > Changing the log level is not the appropriate solution. Make it a<br /> > > configuration parameter.<br/> ><br /> > Why is it inappropriate solution? There's a log_checkpoints GUC that<br /> > drives itand you can either get basic info (summary of the checkpoint) or<br /> > detailed log (with a lower log level).<br />><br /> > In the first patch I've proposed a new GUC (used to set how often the info<br /> > should be loggedor disable it), but Josh Berkus pointed out that I should<br /> > get rid of it if I can. Which is what I've donein the following patches.<br /><p>Well, josh doesn't speak for everybody ;-) <p>Maybe one way could be to change log_checkpointsinto an enum of "off, on, debug "(values open for bikeshedding of course) <p>/Magnus
В списке pgsql-hackers по дате отправления: