[20150326]监测系统参数变化.txt

--转载:
http://carlos-sierra.net/2015/03/25/discovering-if-a-system-level-parameter-has-changed-its-value-and-when-it-happened/

--如果系统参数变化导致性能变化,要修改有时候仅仅通过查看alert*.log文件.
--实际上也可以利用dba_hist_parameter视图,脚本如下:

$ cat parm_mods1.sql
column value format a35
column prior_value format a35

WITH
all_parameters AS (
SELECT snap_id,
       dbid,
       instance_number,
       parameter_name,
       value,
       isdefault,
       ismodified,
       lag(value) OVER (PARTITION BY dbid, instance_number, parameter_hash ORDER BY snap_id) prior_value
  FROM dba_hist_parameter
)
SELECT TO_CHAR(s.begin_interval_time, 'YYYY-MM-DD HH24:MI') begin_time,
       TO_CHAR(s.end_interval_time, 'YYYY-MM-DD HH24:MI') end_time,
       p.snap_id,
       p.dbid,
       p.instance_number,
       p.parameter_name,
       p.value,
       p.isdefault,
       p.ismodified,
       p.prior_value
  FROM all_parameters p,
       dba_hist_snapshot s
WHERE p.value != p.prior_value
   AND s.snap_id = p.snap_id
   AND s.dbid = p.dbid
   AND s.instance_number = p.instance_number
ORDER BY
       s.begin_interval_time DESC,
       p.dbid,
       p.instance_number,
       p.parameter_name
/

--我还有1个脚本来自Kerry Osborne.

$ cat parm_mods.sql
-- parm_mods.sql
--
-- Shows all parameters (including hidden) that have been modified.
-- Uses the lag function so that a single record is returned for each change.
-- It uses AWR data - so only snapshots still in the database will be included.
--
-- The script prompts for a parameter name (which can be wild carded).
-- Leaving the parameter name blank matches any parameter (i.e. it will show all changes).
-- Calculated hidden parameters (those that start with two underscores like "__shared_pool_size")
-- will not be displayed unless requested with a Y.
--
-- Kerry Osborne
--
-- Note: I got this idea from Jeff White.
--
set linesize 155
col time for a15
col parameter_name format a50
col old_value format a30
col new_value format a30
break on instance skip 3
select instance_number instance, snap_id, time, parameter_name, old_value, new_value from (
select a.snap_id,to_char(end_interval_time,'DD-MON-YY HH24:MI') TIME,  a.instance_number, parameter_name, value new_value,
lag(parameter_name,1) over (partition by parameter_name, a.instance_number order by a.snap_id) old_pname,
lag(value,1) over (partition by parameter_name, a.instance_number  order by a.snap_id) old_value ,
decode(substr(parameter_name,1,2),'__',2,1) calc_flag
from dba_hist_parameter a, dba_Hist_snapshot b , v$instance v
where a.snap_id=b.snap_id
and a.instance_number=b.instance_number
and parameter_name like nvl('&parameter_name',parameter_name)
and a.instance_number like nvl('&instance_number',v.instance_number)
)
where
new_value != old_value
and calc_flag not in (decode('&show_calculated','Y',3,2))
order by 1,2
/

--可以使用他们了解修改了那些参数,当然这个受awr保存事件的限制.