Re: Very Strange Migrating Rows experience (Chained Rows)

Can you be a little more precise about how you copy and delete them ?

e.g. copy out into a different table, then commit, delete then commit. Or copy out, delete copy back commit etc.

Do the same PKs end up with their original rowids, or do you get completely new rowids, or the same set of rowids with PKs in different places.

Are these rows the ONLY 140 rows in the table, AVG_ROW_LEN applies to the whole table, how have you determined the actually lengths of these rows ?

Which version of Oracle ?

--

Jonathan Lewis
Yet another Oracle-related web site: http://www.jlcomp.demon.co.uk

Kevin A Lewis wrote in message
<8a%h4.2331$hr4.54539_at_newreader.ukcore.bt.net>...
>I have been trying to remove migrating rows from some tables in a
production
>database and the problem appears to be getting worse rather than better.
>
>I have 140 rows with an AVG_ROW_LEN of 226 in a database of 2k block size
>the table has a PCTFREE of 10 and PCTUSED of 40 (the defaults). All 140 are
>recorded by ANALYZE as CHAINED ROWS. I copy them, delete the original rows
>and write them back; yet on reANALYZE they are recorded as still CHAINED
>ROWS.
>
>Surely they cannot be real chained rows, they are too short. And yet surely
>they cannot be migrated rows as they have just been inserted and suffered
no
>updates at all.
>
>Confused of Ipswich England
>
>Regards
>
>
>
>--
>Kevin A Lewis (BOCM PAULS LTD) - Animal Feed Manufacturer - Ipswich United
>Kingdom)
> <Kevin_A_Lewis_at_Hotmail.com>
>
>The views expressed herein by the author of this document
>are not necessarily those of BOCM PAULS Ltd.
>
>
Received on Fri Jan 21 2000 - 15:16:11 CST