Re: Source for tim=nnnnn in 10046 trace

From: Yong Huang <yong321_at_yahoo.com>
Date: Mon, 26 Aug 2013 07:54:59 -0700 (PDT)
Message-ID: <1377528899.26917.YahooMailNeo_at_web184801.mail.gq1.yahoo.com>

One addition to what I posted earlier (shown below). The derived_sysdate in the query is 1969-12-31 19:59:59, i.e. 8pm instead of 23:59:59 of that night (or 12am of 1970-01-01). The 4 hour difference may be due to different timezones. According to http://www.timeanddate.com/worldclock/converted.html?iso700101T00&p1=0&p29&p3#&p4(3&p5(6

8pm of 1969-12-31 at UTC epoch is the time for the timezone where Nova Scotia, Canada is. I'm not sure why.

Yong Huang

My previous message:

You just need to change your sysdate to 2013-08-23 15:24:25.823. For example:

SQL> with x as (select 1377285865823053 as t, (1377285865823053/1e6) as t1 from dual)
  2  select t tim_micro, t1 tim_sec, '2013-08-23 15:24:25' curr_sysdate, to_date('2013-08-23 15:24:25','yyyy-mm-dd hh24:mi:ss')-(t1/86400)
  3  derived_sysdate
  4  from x;

           TIM_MICRO              TIM_SEC CURR_SYSDATE        DERIVED_SYSDATE

-------------------- -------------------- ------------------- -------------------

    1377285865823053    1377285865.823053 2013-08-23 15:24:25 1969-12-31 19:59:59

The derived sysdate is the epoch.

--
http://www.freelists.org/webpage/oracle-l

Received on Mon Aug 26 2013 - 16:54:59 CEST