Re: DBMS_MONITOR

From: Mladen Gogala <gogala.mladen_at_gmail.com>
Date: Sat, 24 Dec 2011 01:02:49 +0000 (UTC)
Message-ID: <pan.2011.12.24.01.02.49_at_gmail.com>



On Fri, 23 Dec 2011 10:57:43 -0800, Charles Hooper wrote:

> I believe that the session with the trace enabled through DBMS_MONITOR
> must execute at least one SQL statement after tracing is enabled for the
> session, before ORADEBUG will report that the trace is enabled.

I figured out what's happening. I was convinced that DBMS_MONITOR sets a flag in the UGA, but it doesn't do that. It sends a message to the affected process, which will set the flag itself, the next time it is activated . This is an example of what Jonathan calls "AST". DBMS_MONITOR uses the same mechanism as Unix signal delivery: it queues the message, probably incurring the famous "rdbms ipc messsage" wait, and the process will set its own flags on the next activation. The next activation will come when a SQL statement is submitted. OK, now I understand the mechanism.

-- 
http://mgogala.byethost5.com
Received on Fri Dec 23 2011 - 19:02:49 CST

Original text of this message