Mailing List Archive


Back to the month index Back to the list index
David J. Hughes (bambi@Hughes.com.au)
Thu, 15 Apr 1999 08:34:32 +1000 (EST) 

Date: Thu, 15 Apr 1999 08:34:32 +1000 (EST)
From: "David J. Hughes" <bambi@Hughes.com.au>
Subject: Re: Query Logging File
Message-ID: <Pine.BSF.3.96.990415083024.353A-100000@fawn.Hughes.com.au>

On Wed, 14 Apr 1999, Warren Killian wrote:

> 14-Apr-1999 07:45:02 nobody UNIX_SOCK hr 41
> SELECT deptname FROM dept WHERE deptID=2
>
> what does the "41" and the end of the first line mean? Other entries in
> the query log have different numbers. Thanks for the help!

The first line is a log header providing time, date, user type info. The
last field is the number of bytes of actual query text that follows the
query heder line. If you read the first line ,then read 41 bytes, your
next read would fall at the start of the next log header.

The reason it's done this way is because the query text itself can (and
often will) contain new line characters etc.

Have a look at misc/parse_query_log for a Lite script that parses the log
format.

Bambi
...