RE: Long Parse Time

  • From: "Ric Van Dyke" <ric.van.dyke@xxxxxxxxxx>
  • To: <ganstadba@xxxxxxxxxxx>, <oracle-l@xxxxxxxxxxxxx>
  • Date: Thu, 7 May 2009 09:20:29 -0500

* Most interesting was that Ric from Hotsos had troubles gleaning why
the parse time was so long from the 10053 trace file.*

Hey Mike!
Well the problem is that the 10053 doesn't have trimming information in
it.  We can make assumptions about how much time has been taken up in
the 10053 based on volume of "stuff" but that is very dangerous.  So I
didn't. 

This ultimately was an issue that I was not working on directly but one
where I was assisting some one, I don't have direct access to the system
so much as I want to do things like dtrace and check stack dumps and
other things, I can't.  

Yes I would really like to dig into this issue and really find the root
cause, however I'm likely not going to have the opportunity to do that.
I'm less then fully satisfied that we "threw something at it" and it
worked without knowing all the gritty details.  But as we all know there
are times where we just have to count our blessings and move on.

Again thanks to all that contributed here! 
 

-----Original Message-----
From: oracle-l-bounce@xxxxxxxxxxxxx
[mailto:oracle-l-bounce@xxxxxxxxxxxxx] On Behalf Of Michael McMullen
Sent: Thursday, May 07, 2009 5:47 AM
To: oracle-l@xxxxxxxxxxxxx
Subject: RE: Long Parse Time

Very interesting thread. Most interesting was that Ric from Hotsos had
troubles gleaning why the parse time was so long from the 10053 trace
file.
I'm not knocking Ric and I'm a firm believer in Method-R and kudos for
him
coming to the list looking for an answer, but the thought of then having
to
delve into stack traces and then pick some profiler aggregator or even
write
my own gives me nightmares and at the end of the day it still comes down
to
dropping a hint in place.

I am just extremely thankful for guys like Tanel/Ric and all the other
oracle gurus for making my job easier. I feel like my dad (small city
veterinarian) who told me there were lots of better/smarter vets than
him
working on making the profession better so he could do a better job in
his
small practice) and that it was his job to keep abreast of what they
were
doing and how he could apply it.

So, now I condense this thread into a note and index it via google and
add
it to my lexicon and continue to try and do my job better with the help
of
everybody out there.

Mike
-----Original Message-----
From: oracle-l-bounce@xxxxxxxxxxxxx
[mailto:oracle-l-bounce@xxxxxxxxxxxxx]
On Behalf Of Tanel Poder
Sent: Thursday, May 07, 2009 7:09 AM
To: oracle-l@xxxxxxxxxxxxx
Subject: RE: Long Parse Time

Again, if parsing takes minutes or tens of seconds of CPU time then
you're
hitting a bug or someone has fiddled with undocumented parameters!
Parsing
should not take that long! 

Yes one way is to try to disable "this and that" in hope that the
problem
goes away, but the SYSTEMATIC approach would be seeing WHERE the parsing
TIME is spent!




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


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


Other related posts: