Oracle FAQ Your Portal to the Oracle Knowledge Grid
HOME | ASK QUESTION | ADD INFO | SEARCH | E-MAIL US
 

Home -> Community -> Mailing Lists -> Oracle-L -> Re: How to identify large transactions

Re: How to identify large transactions

From: Jonathan Lewis <jonathan_at_jlcomp.demon.co.uk>
Date: Tue, 4 Jan 2005 15:06:44 -0000
Message-ID: <02df01c4f26f$012c7940$6702a8c0@Primary>

Look at v$transaction -

    used_ublk, used_urec undo blocks and records used for updates that are not array-based, each record is roughly equivalent to one change to a table or index block. For example, if you update one row, changing non-null columns that are in three indexes, you will get 7 undo records:

    One for the table, two for each index.

The view also has

    start_time
to tell you when the transaction started, and

    ses_addr
to get you back to the owning session (v$session.saddr)

Regards

Jonathan Lewis

http://www.jlcomp.demon.co.uk/faq/ind_faq.html The Co-operative Oracle Users' FAQ

http://www.jlcomp.demon.co.uk/seminar.html Public Appearances - schedule updated Dec 23rd 2004

I believe that some developers sql is generating large transactions possibly due to lack of commit statements.

How can I identify these large transactions and the sql responsible?

--

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

--

http://www.freelists.org/webpage/oracle-l Received on Tue Jan 04 2005 - 09:02:04 CST

Original text of this message

HOME | ASK QUESTION | ADD INFO | SEARCH | E-MAIL US