Re: audit table containing Select statements submitted
От | Hogan, James F. Jr. |
---|---|
Тема | Re: audit table containing Select statements submitted |
Дата | |
Msg-id | D73D467FE58F004E9D3FA25414945DC1051B99DF@AUSEX3VS1.seton.org обсуждение исходный текст |
Ответ на | audit table containing Select statements submitted ("Hogan, James F. Jr." <JHogan@seton.org>) |
Ответы |
Re: audit table containing Select statements submitted
|
Список | pgsql-hackers |
Thank you all for the effort you put into response. The biggest thing I want to avoid isn't so much having to parse through the log files but to avoid turning on such extensive logging altogether. I am not sure what kind of additional load logging to this extent may add. Looks like I am not going to have much in the way of alternative. Maybe some day. Good news is that most access is via Web Interface and I capture most activity that way. Again thank you all. Jim -----Original Message----- From: Tom Lane [mailto:tgl@sss.pgh.pa.us] Sent: Friday, May 12, 2006 8:03 PM To: Joshua D. Drake Cc: josh@agliodbs.com; Andrew Dunstan; Hogan, James F. Jr.; pgsql-hackers@postgresql.org Subject: Re: [HACKERS] audit table containing Select statements submitted "Joshua D. Drake" <jd@commandprompt.com> writes: > Hmmm... well why don't we add log_line_suffix :) Doesn't help; you'd also need logic in there to quote any XML tags appearing in the message text. At that point, adding a "log_line_suffix" is a transparent pretense of generality --- what you might as well do is just have a full-fledged "emit the log in XML" switch. (I concur with Andrew's comments that this is pretty silly, unless someone wants to go to the further work of XML-ifying the message contents to some reasonable extent. If you are going to have to write a parser to make sense of the message contents, it is really pretty lame to claim that you can't cope with parsing the current log format as-is.) regards, tom lane
В списке pgsql-hackers по дате отправления: