oracleHu的专栏

statspack dump Top 5 Wait Events ~~~~~~~~~~~~~~~~~                                             Wait % Total Event                                               Waits  Time (cs) Wt Time ---------------

create or replace procedure DeleteBigTable(p_tablename       in    varchar2,p_condition       in    varchar2,p_count        in    varchar2)aspragma autonomous_transaction;n_delete number:=0;beginwhile

autotrace 是DBA 对SQL分析的必要工具它的启用必须具备两个要素plustrace职责 -- 查看状态视图的权限plan_table表--存储plan的表plustraceplustrace.sql 建立，分配给要使用autotrace的用户@plustracegrant plustrace to testplan_tableutlplan.sql 建立 share to user w

fnd_request.submit_request你把这个放到数据库包里面maxHu 2005-12-29

1.sql_tracea. 全局在 pfile/spfile 中使用 sql_trace=true导致所有进程的活动都被跟踪，后台进程和用户SESSION ，导致严重的性能问题。慎重使用b.当前SESSION alter session set sql_trace=true ;alter session set sql_trace=flase ;c.跟踪指定的SESSION DBMS_SYSTEM

Xplain.sql1. download it 点击下载此文件2.put it to server3.replace the sql statement to III part of the Xplain.sql4.login sqlplus using owner of scheme5.run @coe_xplain.sql6.read the reasult file coe_statmen

Optimizer_mode : RULE | FRIST_ROWS | ALL_ROWS | CHOOSECOST : run by FTS Which dont need , Iespeacial under 3 tables jionRULE : choose error index FULL TABLE SCAN 临界条件:1. sorted table -- > 40% rows ne

Table rebuild1: CATS         ORDER BY  (best solution)              create table new_customer parallel (drgree 11) as select * from customer oder by customer_number         INDEX TIPS               cr

1: oracle unveral install --> add oracle advance network option 2: modify sqlnet.ora add follow lines  at both server side and client side# SQLNET.ORA Network Configuration File: h:/oracle/ora92/NETWO

locks and locked objects ============================== select /*+RULE*/  s.machine,  s.osuser "O/S-User/",  s.username "Ora-User/",  s.sid "Session-ID/",  s.serial# "Serial/",  s.process "Process-ID/

0.Read readme document1.applmgr用户登录（orafinc之流）2.设置环境变量（ $APPL_TOP/APPSORA.env）3.停应用 adstpall.sh （至少要停并发服务器）,数据库和监听保持启动状态 （其实大多数都无所谓）4.apply english patch:cxxx.drv ,dxxx.drv,gxxx.drv 命令行：adpatch5.apply

一般情况下数据文件出现block corrugation ，正常的恢复方法是使用备份。如果没有备份，可采用以下方法补救。1.使用 dbv工具 和dba_exents试图定位corrugate block及其表2.ALTER SYSTEM SET EVENTS=10231 trace name context forever,level 10 ; (内部事件，设置在全表扫描时跳过