Re: sysaux purge problems

From: Jeffrey Beckstrom <JBECKSTROM_at_gcrta.org>
Date: Fri, 05 Sep 2014 13:49:29 -0400
Message-Id: <5409BF690200000B0052C5E2_at_groupwise.gcrta.org>

Was this database clone? If so, you may have more than one DBID in the table. In our systems, it seems that the "source" database stays in the table and is not purged.

>>> Jeremy Schneider <jeremy.schneider_at_ardentperf.com> 9/5/14 1:36 PM >>>
I've reviewed the various MOS docs and bugs related to sysaux purge problems already. I'm just wondering if anyone else out there has specifically seen inconsistencies where it seems that the AWR is purging fine except for only the ASH table? Or where DBMS_STATS.GET_STATS_HISTORY_AVAILABILITY reports that data has been purged, but a select on dba_tab_stats_history shows otherwise?

maybe one of the existing bugs causes this behavior, just seemed to me from the descriptions that get_stats_history_availability is generally accurate. i'll work the SR and everything, just wondered if this problem was already solved and I just missed the doc. the initial support guys aren't bad, but sometimes oracle-l buddies correct my search misses faster. :)

btw this is on 11.2.0.3 psu11.

• AWR/ASH: SQL> select min(begin_interval_time) from dba_hist_snapshot; MIN(BEGIN_INTERVAL_TIME)

  ---

  2014-05-23 01:30:11

SQL> select min(sample_time) from DBA_HIST_ACTIVE_SESS_HISTORY; MIN(SAMPLE_TIME)

---

2013-10-09 02:00:01

• stats history: SQL> SELECT DBMS_STATS.GET_STATS_HISTORY_AVAILABILITY FROM DUAL; GET_STATS_HISTORY_AVAILABILITY

  ---

  2014-08-05 00:01:08

SQL> select min(STATS_UPDATE_TIME) from dba_tab_stats_history; MIN(STATS_UPDATE_TIME)

---

2013-09-15 01:01:06

SQL> select min(savtime) from WRI$_OPTSTAT_HISTHEAD_HISTORY; MIN(SAVTIME)

---

2013-08-04 21:19:35

--
http://about.me/jeremy_schneider



.



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

Received on Fri Sep 05 2014 - 19:49:29 CEST