http://blog.csdn.net/tianlesoftware/article/details/6319198
      Oracle AWR 报告能很好的提供有关DB性能的信息。 所以DBA 需要定期的查看AWR的报告。 有关AWR报告的说明参考:

       Oracle AWR 介绍

       http://blog.csdn.net/tianlesoftware/archive/2009/10/17/4682300.aspx

 

       如果说每天都去手工的出创建AWR报告,是个费时的活。 所以,最好就是弄成脚本来执行。 这篇文章就是说明这个问题的。

 

       注意的一个问题,AWR 报告的两个snap 之间不能有重启DB的操作。

 

Python 执行 系统命令

http://blog.csdn.net/tianlesoftware/archive/2011/02/17/6192202.aspx

 

Python 发送带 附件 邮件 脚本

http://blog.csdn.net/tianlesoftware/archive/2011/04/12/6318199.aspx

 

 

. 准备工作

 

       一般我们都是条用awrrpt.sql 来创建我们的AWR报告。 我们先看下这个脚本的具体内容:

[oracle@rac1 admin]$ cat awrrpt.sql | grep -v 'Rem'|grep -v '^--'

set echo off heading on underline on;

column inst_num  heading "Inst Num"  new_value inst_num  format 99999;

column inst_name heading "Instance"  new_value inst_name format a12;

column db_name   heading "DB Name"   new_value db_name   format a12;

column dbid      heading "DB Id"     new_value dbid      format 9999999999 just c;

 

prompt

prompt Current Instance

prompt ~~~~~~~~~~~~~~~~

 

select d.dbid            dbid

     , d.name            db_name

     , i.instance_number inst_num

     , i.instance_name   inst_name

  from v$database d,

       v$instance i;

 

@@awrrpti

 

undefine num_days;

undefine report_type;

undefine report_name;

undefine begin_snap;

undefine end_snap;

 

       在以上的脚本里,我们发现它只是生成了一些变量,然后把这些变量传给了另一个脚本:awrrpti.sql 我们看下awrrpti.sql 脚本的具体内容:

 

[oracle@rac1 admin]$ cat awrrpti.sql | grep -v 'Rem'|grep -v '^--' 

 

set echo off;

set veri off;

set feedback off;

 

variable rpt_options number;

 

define NO_OPTIONS   = 0;

define ENABLE_ADDM  = 8;

 

 

begin

  :rpt_options := &NO_OPTIONS;

end;

/

 

prompt

prompt Specify the Report Type

prompt ~~~~~~~~~~~~~~~~~~~~~~~

prompt Would you like an HTML report, or a plain text report?

prompt Enter 'html' for an HTML report, or 'text' for plain text

prompt  Defaults to 'html'

 

column report_type new_value report_type;

set heading off;

select 'Type Specified: ',lower(nvl('&&report_type','html')) report_type from dual;

set heading on;

 

set termout off;

column ext new_value ext;

select '.html' ext from dual where lower('&&report_type') <> 'text';

select '.txt' ext from dual where lower('&&report_type') = 'text';

set termout on;

 

@@awrinput.sql

-- 这个脚本主要是确定SNAP的。

@@awrinpnm.sql 'awrrpt_' &&ext

-- 这个脚本主要是确定AWR 文件名称的

 

set termout off;

column fn_name new_value fn_name noprint;

select 'awr_report_text' fn_name from dual where lower('&report_type') = 'text';

select 'awr_report_html' fn_name from dual where lower('&report_type') <> 'text';

 

column lnsz new_value lnsz noprint;

select '80' lnsz from dual where lower('&report_type') = 'text';

select '1500' lnsz from dual where lower('&report_type') <> 'text';

 

set linesize &lnsz;

set termout on;

spool &report_name;

 

select output from table(dbms_workload_repository.&fn_name( :dbid,

                                                    :inst_num,

                                                    :bid, :eid,

                                                    :rpt_options ));

 

 

spool off;

 

prompt Report written to &report_name.

 

set termout off;

clear columns sql;

ttitle off;

btitle off;

repfooter off;

set linesize 78 termout on feedback 6 heading on;

undefine report_name

 

undefine report_type

undefine ext

undefine fn_name

undefine lnsz

 

undefine NO_OPTIONS

undefine ENABLE_ADDM

 

undefine top_n_events

undefine num_days

undefine top_n_sql

undefine top_pct_sql

undefine sh_mem_threshold

undefine top_n_segstat

 

whenever sqlerror continue;

[oracle@rac1 admin]$

 

 

       这个脚本才是我们真正生成AWR的脚本。 在这个脚本里面,提示我们选择AWR报告的类型。

 

 

通过上面的2个脚本,我们将AWR报告简化一下:

       select output from

table(dbms_workload_repository.&fn_name(:dbid, :inst_num,:bid, :eid,:rpt_options ));

 

这条语句就是整个AWR报告的核心:

1&fn_name 决定AWR报告的类型,有2个值:awr_report_htmlawr_report_text

2dbidinst_num,bid,eid 可以通过dba_hist_snapshot查询. bid 指的是begin snap_id, eid 指的是end snap_id.

 

 

SQL> select * from (select snap_id,dbid,instance_number from dba_hist_snapshot  order by snap_id) where rownum<10;

 

   SNAP_ID       DBID INSTANCE_NUMBER

---------- ---------- ---------------

       184  809910293               2

       184  809910293               1

       185  809910293               2

       185  809910293               1

       186  809910293               2

       186  809910293               1

       187  809910293               2

       187  809910293               1

       188  809910293               2

 

9 rows selected.

 

       我这里是个RAC 环境, 通过这个可以看出在每个节点上都保存着AWR的信息。

 

3rpt_options:该参数控制是否显示ADDM的。

--  NO_OPTIONS -

--    No options. Setting this will not show the ADDM

--    specific portions of the report.

--    This is the default setting.

--

--  ENABLE_ADDM -

--    Show the ADDM specific portions of the report.

--    These sections include the Buffer Pool Advice,

--    Shared Pool Advice, PGA Target Advice, and

--    Wait Class sections.

define NO_OPTIONS   = 0;

define ENABLE_ADDM  = 8;

 

 

有了上面的数据之后,我们就可以使用如下SQL直接生成AWR报告了。

SQL>select output from table(dbms_workload_repository.awr_report_html(809910293, 2,220,230,0));

 

SQL>select output from table(dbms_workload_repository.awr_report_text(809910293, 2,220,230,0));

 

 

. 生成AWR报告 SQL脚本

       以上写了这么多,就是为了一个脚本:myawrrpt.sql.  这个脚本就是自动的去收集信息。 因为如果我们是调用awrrpt.sql的话,需要输入一些参数。 我们修改一下脚本,让它根据我们的需求来收集信息,这样就不用输入参数了。

 

[oracle@rac1 admin]$ cat myawrrpt.sql

conn / as sysdba;

set echo off;

set veri off;

set feedback off;

set termout on;

set heading off;

 

variable rpt_options number;

 

define NO_OPTIONS = 0;

define ENABLE_ADDM = 8;

 

-- according to your needs, the value can be 'text' or 'html'

define report_type='html';

begin

:rpt_options := &NO_OPTIONS;

end;

/

 

variable dbid number;

variable inst_num number;

variable bid number;

variable eid number;

begin

select max(snap_id)-48 into :bid from dba_hist_snapshot;

select max(snap_id) into :eid from dba_hist_snapshot;

select dbid into :dbid from v$database;

select instance_number into :inst_num from v$instance;

end;

/

 

column ext new_value ext noprint

column fn_name new_value fn_name noprint;

column lnsz new_value lnsz noprint;

 

--select 'txt' ext from dual where lower('&report_type') = 'text';

select 'html' ext from dual where lower('&report_type') = 'html';

--select 'awr_report_text' fn_name from dual where lower('&report_type') = 'text';

select 'awr_report_html' fn_name from dual where lower('&report_type') = 'html';

--select '80' lnsz from dual where lower('&report_type') = 'text';

select '1500' lnsz from dual where lower('&report_type') = 'html';

 

set linesize &lnsz;

 

-- print the AWR results into the report_name file using the spool command:

 

column report_name new_value report_name noprint;

select 'awr'||'.'||'&ext' report_name from dual;

set termout off;

spool &report_name;

select output from table(dbms_workload_repository.&fn_name(:dbid, :inst_num,:bid, :eid,:rpt_options ));

spool off;

 

 

set termout on;

clear columns sql;

ttitle off;

btitle off;

repfooter off;

undefine report_name

undefine report_type

undefine fn_name

undefine lnsz

undefine NO_OPTIONS

exit

[oracle@rac1 admin]$

 

       这个脚本是收集过去48个小时的snap 来生成AWR 生成的文件名称是awr .html,这个也是spool 指定的,可以生成其他名称。

 

 

. 自动上传AWRPython脚本

 

       在这个脚本里做2件事,第一是调用第二步里的SQL脚本,生成awr报告,然后将AWR 发送到指定邮箱。

 

 

createSendAWR.py

 

#!/usr/bin/python

#coding=gbk

#created by tianlesoftware

#2011-4-12

 

import os

import sys

import smtplib

import pickle

import mimetypes

from email.MIMEText import MIMEText

from email.MIMEImage import MIMEImage

from email.MIMEMultipart import MIMEMultipart

 

 

SMTP_SERVER='192.168.1.120'

EMAIL_USER='user'

EMAIL_PASSWD='pwd'

EMAIL_SUBJECT='192.168.88.209 AWR Report'

FROM_USER='[email protected]'

TO_USERS=['[email protected]','[email protected]']

 

 

def createawr():

  pipe = os.popen(' /u01/app/oracle/product/10.2.0/db_1/bin/sqlplus /nolog @awrrpt.sql')

 

def mysendmail(fromaddr,toaddrs,subject):

 

        COMMASPACE=','

        msg = MIMEMultipart()

        msg['From'] = fromaddr

        msg['To'] = COMMASPACE.join(toaddrs)

        msg['Subject'] = subject

 

     

        txt = MIMEText("192.168.88.209 AWR Report, The report be send at 9 AM every day ")

        msg.attach(txt)  

 

  

        fileName = r'/home/oracle/awr.html'

        ctype, encoding = mimetypes.guess_type(fileName)

        if ctype is None or encoding is not None:

            ctype = 'application/octet-stream'

        maintype, subtype = ctype.split('/', 1)

        att = MIMEImage((lambda f: (f.read(), f.close()))(open(fileName, 'rb'))[0], _subtype = subtype)

        att.add_header('Content-Disposition', 'attachment', filename = fileName)

        msg.attach(att)  

 

        server=smtplib.SMTP(SMTP_SERVER)

        server.login(EMAIL_USER,EMAIL_PASSWD)

        server.sendmail(fromaddr,toaddrs,msg.as_string())

        server.quit()

 

if __name__=='__main__':

               createawr()

        mysendmail(FROM_USER, TO_USERS, EMAIL_SUBJECT)

        #print 'send successful'

 

 

. Python 添加到crontab

 

 

[oracle@qs-wg-db1 scripts]$ crontab -l

 

40 17 * * * export ORACLE_HOME='/home/oracle_app' && ORACLE_SID=XX  && cd /u01/backup/scripts && /u01/backup/scripts/createSendAWR.py>/u01/backup/scripts/createSendAWR.log 2>&1

 

我这里因为报了

SP2-0750: You may need to set ORACLE_HOME to your Oracle software directory

的错误,所以把变量加了上去。

 

 

  Linux Crontab 定时任务 命令详解

  http://blog.csdn.net/tianlesoftware/archive/2010/02/21/5315039.aspx

10-05 02:19