Re: logging blemishes
От | Bruce Momjian |
---|---|
Тема | Re: logging blemishes |
Дата | |
Msg-id | 200509220311.j8M3BNR02439@candle.pha.pa.us обсуждение исходный текст |
Ответ на | Re: logging blemishes (Andrew Dunstan <andrew@dunslane.net>) |
Ответы |
Re: logging blemishes
|
Список | pgsql-hackers |
Andrew Dunstan wrote: > > > Bruce Momjian wrote: > > >Andrew Dunstan wrote: > > > > > >>>>Example (log_line_prefix = '%t %q%u@%d %r %p %c:%l'): > >>>> > >>>>2005-09-19 19:16:39 EDT [unknown]@[unknown] 6541 432f46d7.198d:1 LOG: > >>>>connection received: host=[local] port= > >>>> > >>>> > >>>> > >>>> > >>> > >>> > >>In the case above, I rather expected %q to kick in. With the additional > >>tests it would. > >> > >>It's debatable, though, and not hugely important either way, I think. > >> > >> > > > >Are you saying "connection received" should honor %q? It seems it is a > >session line, rather than a server line, no? > > > > > > > > Maybe, the line just struck me as rather ugly. Never mind. Now that you mention it, the log line for connections does look wrong. "[local]" doesn't have remote port numbers like tcp does. The remote port is the same number as the server port. I am thinking we should suppress the 'port=' output for local connections. We properly suppress the port number in parentheses for log_line_prefix='%r'. -- Bruce Momjian | http://candle.pha.pa.us pgman@candle.pha.pa.us | (610) 359-1001+ If your life is a hard drive, | 13 Roberts Road + Christ can be your backup. | Newtown Square, Pennsylvania19073
В списке pgsql-hackers по дате отправления: