logging changes - sample output - Mailing list pgsql-hackers

From Andrew Dunstan
Subject logging changes - sample output
Date
Msg-id 3F354143.1050603@dunslane.net
Whole thread Raw
List pgsql-hackers
I ended up doing more or less what Bruce suggested, for 2 reasons - 
first it is extensible, and second it avoids hardcoding any special 
separators.

The new config variable is set like this:
   log_line_format = '<%U%%%D> '     # %U=username %D=databasename %%=%

Sample output including session logging when log_connections= true:

2003-08-09 14:33:22 [14219] LOG:  database system was shut down at 
2003-08-09 14:33:17 EDT
2003-08-09 14:33:22 [14219] LOG:  checkpoint record is at 0/27A3470
2003-08-09 14:33:22 [14219] LOG:  redo record is at 0/27A3470; undo 
record is at 0/0; shutdown TRUE
2003-08-09 14:33:22 [14219] LOG:  next transaction id: 6235; next oid: 
154553
2003-08-09 14:33:22 [14219] LOG:  database system is ready
2003-08-09 14:33:35 [14222] LOG:  connection received: host=localhost port=
2003-08-09 14:33:35 [14222] <andrew%adtest> LOG:  connection authorized: 
user=andrew database=adtest
2003-08-09 14:33:53 [14222] <andrew%adtest> LOG:  query: create table 
foo (x text, y text);
2003-08-09 14:34:07 [14222] <andrew%adtest> LOG:  session ended: 
duration: 0:00:32.807689 user=andrew database=adtest
2003-08-09 14:34:22 [14217] LOG:  received fast shutdown request
2003-08-09 14:34:22 [14225] LOG:  shutting down
2003-08-09 14:34:25 [14225] LOG:  database system is shut down


I'll clean this up (don't think we really need microsec granularity on 
session logging) and submit the patch this weekend.

cheers

andrew




pgsql-hackers by date:

Previous
From: Andreas Pflug
Date:
Subject: Re: getting confused parsing ACLITEMS...
Next
From: Andreas Pflug
Date:
Subject: Re: Batch Operations