Re: consistent read and LIO question

From: lsllcm <lsllcm_at_gmail.com>
Date: Fri, 15 May 2009 09:26:24 -0700 (PDT)
Message-ID: <b4e8b82e-f898-40ef-ac19-ba467a5bf85c_at_a5g2000pre.googlegroups.com>



On May 15, 5:13 pm, mathewbutler <mathewbut..._at_yahoo.com> wrote:
> If I read your trace correctly it shows 5 consistent reads taking
> place, but two of these didn't find the required blocks in the cache
> and therefore had to read from disk via 'db file sequential read'.
>
> If so, it looks like v$sysstat only records LIO that do not initiate a
> physical read.
>
> Mathew Butler

Hi Mathew,

Yes, 5 consistent reads is right. I use v$sysstat and v$statname wrongly. The correct way is to run the following in other session,

-----------------------------------begin-------------------------
*** In other session, run the following (where SID = the sid of the other session) before and after an execution of the select statement in the other session.
select n.name, s.value from v$sesstat s, v$statname n where s.statistic# = n.statistic# and s.sid = 123 and n.name like 'consistent
%';
-----------------------------------end-------------------------

I got it from below article

http://richardfoote.wordpress.com/2007/12/21/differences-between-unique-and-non-unique-indexes-part-ii/

Thanks
Jacky Received on Fri May 15 2009 - 11:26:24 CDT

Original text of this message