RE: Chained vs. migrated rows - Any easy way to tell the difference?

From: <Joel.Patterson_at_crowley.com>
Date: Mon, 3 Nov 2008 08:14:02 -0500
Message-ID: <0684DA55864E404F8AD2E2EBDFD557DA01E30897@JAXMSG01.crowley.com>


Thanks, I believe is was a long lost recollection that was re-introduced by Mark leading to a new chained_rows table that can handle IOT's. That fixed the 01495 error anyway. The difference between dba_tables.chain_cnt and what is contained in chained_rows table now is still a mystery. Stats are collected nightly.

Joel Patterson
Database Administrator
joel.patterson_at_crowley.com
x72546
904 727-2546

-----Original Message-----

From: Yong Huang [mailto:yong321_at_yahoo.com] Sent: Saturday, November 01, 2008 9:17 PM To: Patterson, Joel
Cc: oracle-l_at_freelists.org
Subject: RE: Chained vs. migrated rows - Any easy way to tell the difference?

Joel,

Regarding "ORA-01495: specified chain row table not found", a systematic way to troubleshoot this type of problem is to enable SQL trace:

alter session set events '10046 trace name context forever, level 4'

followed by your analyze list chained rows. The trace file will show you exactly where it goes wrong. If the trace file doesn't show or doesn't show enough recursive SQLs, flush shared pool and try again. If you absolutely cannot flush shared pool, try some "harmless" DDL on the table and recreate the chained_rows table that you think is used to get the recursive SQLs out of dictionary cache.

Yong Huang       

--

http://www.freelists.org/webpage/oracle-l Received on Mon Nov 03 2008 - 07:14:02 CST

Original text of this message