oracle常用命令收藏

1、查看表空间的名称及大小

 

select t.tablespace_name,round(sum(bytes/(1024*1024)),0) ts_size

 

from dba_tablespaces t, dba_data_files d

 

where t.tablespace_name = d.tablespace_name

 

group by t.tablespace_name;

 

2、查看表空间物理文件的名称及大小

 

select tablespace_name, file_id, file_name,

 

round(bytes/(1024*1024),0) total_space

 

from dba_data_files

 

order by tablespace_name;

 

3、查看回滚段名称及大小

 

 

 

select segment_name, tablespace_name,r.status,

 

(initial_extent/1024)InitialExtent,(next_extent/1024) NextExtent,

 

max_extents, v.curext CurExtent

 

From dba_rollback_segs r, v$rollstat v

 

Where r.segment_id = v.usn(+)

 

order by segment_name;

 

 

 

4、查看控制文件

 

 

 

select name from v$controlfile;

 

 

 

5、查看日志文件

 

 

 

select member from v$logfile;

 

 

 

6、查看表空间的使用情况

 

 

 

select sum(bytes)/(1024*1024) asfree_space,tablespace_name

 

from dba_free_space

 

group by tablespace_name;

 

 

 

SELECT A.TABLESPACE_NAME,A.BYTESTOTAL,B.BYTES USED, C.BYTES FREE,

 

(B.BYTES*100)/A.BYTES "%USED",(C.BYTES*100)/A.BYTES "% FREE"

 

FROM SYS.SM$TS_AVAIL A,SYS.SM$TS_USEDB,SYS.SM$TS_FREE C

 

WHERE A.TABLESPACE_NAME=B.TABLESPACE_NAMEAND A.TABLESPACE_NAME=C.TABLESPACE_NAME;

 

 

 

7、查看数据库库对象

 

 

 

select owner, object_type, status, count(*)count# from all_objects group by owner, object_type, status;

 

 

 

8、查看数据库的版本　

 

 

 

Select version FROMProduct_component_version

 

Where SUBSTR(PRODUCT,1,6)='Oracle';

 

 

 

9、查看数据库的创建日期和归档方式

 

 

 

Select Created, Log_Mode, Log_Mode FromV$Database;

 

 

 

10、捕捉运行很久的SQL

 

 

 

column username format a12

 

column opname format a16

 

column progress format a8

 

 

 

select username,sid,opname,

 

     round(sofar*100 / totalwork,0) || '%' as progress,

 

     time_remaining,sql_text

 

from v$session_longops , v$sql

 

where time_remaining <> 0

 

and sql_address = address

 

and sql_hash_value = hash_value

 

/

 

11。查看数据表的参数信息

 

SELECT  partition_name, high_value, high_value_length, tablespace_name,

 

       pct_free, pct_used, ini_trans, max_trans, initial_extent,

 

       next_extent, min_extent, max_extent, pct_increase, FREELISTS,

 

       freelist_groups, LOGGING, BUFFER_POOL, num_rows, blocks,

 

       empty_blocks, avg_space, chain_cnt, avg_row_len, sample_size,

 

       last_analyzed

 

  FROM dba_tab_partitions

 

 --WHERE table_name = :tname AND table_owner = :towner

 

ORDER BY partition_position

 

 

 

12.查看还没提交的事务

 

select * from v$locked_object;

 

select * from v$transaction;

 

 

 

13。查找object为哪些进程所用

 

select

 

p.spid,

 

s.sid,

 

s.serial# serial_num,

 

s.username user_name,

 

a.type object_type,

 

s.osuser os_user_name,

 

a.owner,

 

a.object object_name,

 

decode(sign(48 - command),

 

1,

 

to_char(command), 'Action Code #' ||to_char(command) ) action,

 

p.program oracle_process,

 

s.terminal terminal,

 

s.program program,

 

s.status session_status  

 

from v$session s, v$access a, v$processp  

 

where s.paddr = p.addr and

 

    s.type = 'USER' and   

 

    a.sid = s.sid   and

 

 a.object='SUBSCRIBER_ATTR'

 

order by s.username, s.osuser

 

 

 

14。回滚段查看

 

select rownum,sys.dba_rollback_segs.segment_name Name, v$rollstat.extents

 

Extents, v$rollstat.rssize Size_in_Bytes,v$rollstat.xacts XActs,

 

v$rollstat.gets Gets, v$rollstat.waitsWaits, v$rollstat.writes Writes,

 

sys.dba_rollback_segs.status status fromv$rollstat, sys.dba_rollback_segs,

 

v$rollname where v$rollname.name(+) =sys.dba_rollback_segs.segment_name and

 

v$rollstat.usn (+) = v$rollname.usn orderby rownum

 

 

 

15。耗资源的进程（top session）

 

select s.schemaname schema_name,    decode(sign(48 - command), 1,

 

to_char(command), 'Action Code #' ||to_char(command) ) action,    status

 

session_status,   s.osuser os_user_name,   s.sid,         p.spid ,         s.serial# serial_num,  

 

nvl(s.username, '[Oracle process]')user_name,   s.terminal terminal,   

 

s.program program,   st.value criteria_value  from v$sesstat st,   v$session s , v$process p  

 

where st.sid = s.sid and   st.statistic# = to_number('38') and   ('ALL' = 'ALL'

 

or s.status = 'ALL') and p.addr = s.paddrorder by st.value desc,  p.spid asc,s.username asc, s.osuser asc

 

 

 

16。查看锁（lock）情况

 

select /*+ RULE */ ls.osuseros_user_name,   ls.usernameuser_name,  

 

decode(ls.type, 'RW', 'Row wait enqueuelock', 'TM', 'DML enqueue lock', 'TX',

 

'Transaction enqueue lock', 'UL', 'Usersupplied lock') lock_type,  

 

o.object_name object,   decode(ls.lmode, 1, null, 2, 'Row Share', 3,

 

'Row Exclusive', 4, 'Share', 5, 'Share RowExclusive', 6, 'Exclusive', null)

 

lock_mode,    o.owner,  ls.sid,   ls.serial#serial_num,   ls.id1,   ls.id2   

 

from sys.dba_objects o, (   select s.osuser,    s.username,    l.type,    

 

l.lmode,   s.sid,    s.serial#,    l.id1,   l.id2   from v$session s,    

 

v$lock l  where s.sid = l.sid ) ls  whereo.object_id = ls.id1 and    o.owner

 

<> 'SYS'   order by o.owner, o.object_name

 

 

 

17。查看等待（wait）情况

 

SELECT v$waitstat.class, v$waitstat.countcount, SUM(v$sysstat.value) sum_value

 

FROM v$waitstat, v$sysstat WHEREv$sysstat.name IN ('db block gets',

 

'consistent gets') group byv$waitstat.class, v$waitstat.count

 

 

 

18。查看sga情况

 

SELECT NAME, BYTES FROM SYS.V_$SGASTATORDER BY NAME ASC

 

 

 

19。查看catched object

 

SELECT owner,              name,              db_link,              namespace, 

 

          type,             sharable_mem,             loads,              executions,  

 

         locks,              pins,              kept        FROM v$db_object_cache

 

         

 

20。查看V$SQLAREA

 

SELECT SQL_TEXT, SHARABLE_MEM,PERSISTENT_MEM, RUNTIME_MEM, SORTS,

 

VERSION_COUNT, LOADED_VERSIONS,OPEN_VERSIONS, USERS_OPENING, EXECUTIONS,

 

USERS_EXECUTING, LOADS, FIRST_LOAD_TIME,INVALIDATIONS, PARSE_CALLS, DISK_READS,

 

BUFFER_GETS, ROWS_PROCESSED FROM V$SQLAREA

 

 

 

21。查看object分类数量

 

select decode (o.type#,1,'INDEX' ,2,'TABLE' , 3 , 'CLUSTER' , 4, 'VIEW' , 5 ,

 

'SYNONYM' , 6 , 'SEQUENCE' , 'OTHER' )object_type , count(*) quantity from

 

sys.obj$ o where o.type# > 1 group bydecode (o.type#,1,'INDEX' , 2,'TABLE' , 3

 

, 'CLUSTER' , 4, 'VIEW' , 5 , 'SYNONYM' , 6, 'SEQUENCE' , 'OTHER' ) union select

 

'COLUMN' , count(*) from sys.col$ unionselect 'DB LINK' , count(*) from

 

 

 

22。按用户查看object种类

 

select u.name schema,   sum(decode(o.type#, 1, 1, NULL))indexes,  

 

sum(decode(o.type#, 2, 1, NULL))tables,   sum(decode(o.type#, 3, 1,NULL))

 

clusters,  sum(decode(o.type#, 4, 1, NULL)) views,  sum(decode(o.type#, 5, 1,

 

NULL)) synonyms,   sum(decode(o.type#, 6, 1, NULL))sequences,  

 

sum(decode(o.type#, 1, NULL, 2, NULL, 3,NULL, 4, NULL, 5, NULL, 6, NULL, 1))

 

others  from sys.obj$ o, sys.user$ u   where o.type# >= 1 and    u.user# =

 

o.owner# and   u.name <> 'PUBLIC'   group by u.name    order by

 

sys.link$ union select 'CONSTRAINT' ,count(*) from sys.con$

 

 

 

23。有关connection的相关信息

 

1）查看有哪些用户连接

 

select s.osuser os_user_name,    decode(sign(48 - command), 1,to_char(command),

 

'Action Code #' || to_char(command) )action,     p.programoracle_process,    

 

status session_status,    s.terminal terminal,    s.program program,   

 

s.username user_name,    s.fixed_table_sequence activity_meter,    '' query,   

 

0 memory,   0 max_memory,     0cpu_usage,    s.sid,   s.serial# serial_num   

 

from v$session s,    v$process p   where s.paddr=p.addr and    s.type = 'USER' 

 

order by s.username, s.osuser

 

2）根据v.sid查看对应连接的资源占用等情况

 

select n.name,

 

 v.value,

 

 n.class,

 

 n.statistic# 

 

from v$statname n,

 

 v$sesstat v

 

where v.sid = 71 and

 

 v.statistic# = n.statistic#

 

order by n.class, n.statistic#

 

3）根据sid查看对应连接正在运行的sql

 

select /*+ PUSH_SUBQ */

 

 command_type,

 

 sql_text,

 

 sharable_mem,

 

 persistent_mem,

 

 runtime_mem,

 

 sorts,

 

 version_count,

 

 loaded_versions,

 

 open_versions,

 

 users_opening,

 

 executions,

 

 users_executing,

 

 loads,

 

 first_load_time,

 

 invalidations,

 

 parse_calls,

 

 disk_reads,

 

 buffer_gets,

 

 rows_processed,

 

 sysdate start_time,

 

 sysdate finish_time,

 

 '>' || address sql_address,

 

 'N'status

 

from v$sqlarea

 

where address = (select sql_address fromv$session where sid = 71)

 

 

 

24．查询表空间使用情况

 

select a.tablespace_name "表空间名称",

 

100-round((nvl(b.bytes_free,0)/a.bytes_alloc)*100,2)"占用率(%)",

 

round(a.bytes_alloc/1024/1024,2) "容量(M)",

 

round(nvl(b.bytes_free,0)/1024/1024,2)"空闲(M)",

 

round((a.bytes_alloc-nvl(b.bytes_free,0))/1024/1024,2)"使用(M)",

 

Largest "最大扩展段(M)",

 

to_char(sysdate,'yyyy-mm-dd hh24:mi:ss')"采样时间"

 

from (select f.tablespace_name,

 

  sum(f.bytes) bytes_alloc,

 

  sum(decode(f.autoextensible,'YES',f.maxbytes,'NO',f.bytes)) maxbytes

 

from dba_data_files f

 

group by tablespace_name) a,

 

(select f.tablespace_name,

 

   sum(f.bytes) bytes_free

 

from dba_free_space f

 

group by tablespace_name) b,

 

(select round(max(ff.length)*16/1024,2)Largest,

 

  ts.name tablespace_name

 

from sys.fet$ ff, sys.file$ tf,sys.ts$ ts

 

where ts.ts#=ff.ts# andff.file#=tf.relfile# and ts.ts#=tf.ts#

 

group by ts.name, tf.blocks) c

 

where a.tablespace_name = b.tablespace_nameand a.tablespace_name = c.tablespace_name

 

 

 

25. 查询表空间的碎片程度

 

 

 

selecttablespace_name,count(tablespace_name) from dba_free_space group bytablespace_name

 

having count(tablespace_name)>10;

 

 

 

alter tablespace name coalesce;

 

alter table name deallocate unused;

 

 

 

create or replace view ts_blocks_v as

 

selecttablespace_name,block_id,bytes,blocks,'free space' segment_name fromdba_free_space

 

union all

 

selecttablespace_name,block_id,bytes,blocks,segment_name from dba_extents;

 

 

 

select * from ts_blocks_v;

 

 

 

select tablespace_name,sum(bytes),max(bytes),count(block_id)from dba_free_space

 

group by tablespace_name;

 

 

 

26。查询有哪些数据库实例在运行

 

select inst_name from v$active_instances;

 

 

 

===========================================================

 

######### 创建数据库----look$ORACLE_HOME/rdbms/admin/buildall.sql #############

 

 

 

create database db01

 

maxlogfiles 10

 

maxdatafiles 1024

 

maxinstances 2

 

logfile

 

GROUP 1('/u01/oradata/db01/log_01_db01.rdo') SIZE 15M,

 

GROUP 2 ('/u01/oradata/db01/log_02_db01.rdo')SIZE 15M,

 

GROUP 3('/u01/oradata/db01/log_03_db01.rdo') SIZE 15M,

 

datafile'u01/oradata/db01/system_01_db01.dbf') SIZE 100M,

 

undo tablespace UNDO

 

datafile'/u01/oradata/db01/undo_01_db01.dbf' SIZE 40M

 

default temporary tablespace TEMP

 

tempfile'/u01/oradata/db01/temp_01_db01.dbf' SIZE 20M

 

extent management local uniform size 128k

 

character set AL32UTE8

 

national character set AL16UTF16

 

set time_zone='America/New_York';

 

 

 

############### 数据字典##########

 

 

 

set wrap off

 

 

 

select * from v$dba_users;

 

 

 

grant select on table_name to user/rule;

 

 

 

select * from user_tables;

 

 

 

select * from all_tables;

 

 

 

select * from dba_tables;

 

 

 

revoke dba from user_name;

 

 

 

shutdown immediate

 

 

 

startup nomount

 

 

 

select * from v$instance;

 

 

 

select * from v$sga;

 

 

 

select * from v$tablespace;

 

 

 

alter session set nls_language=american;

 

 

 

alter database mount;

 

 

 

select * from v$database;

 

 

 

alter database open;

 

 

 

desc dictionary

 

 

 

select * from dict;

 

 

 

desc v$fixed_table;

 

 

 

select * from v$fixed_table;

 

 

 

set oracle_sid=foxconn

 

 

 

select * from dba_objects;

 

 

 

set serveroutput on

 

 

 

execute dbms_output.put_line('sfasd');

 

 

 

############# 控制文件###########

 

 

 

select * from v$database;

 

 

 

select * from v$tablespace;

 

 

 

select * from v$logfile;

 

 

 

select * from v$log;

 

 

 

select * from v$backup;

 

 

 

/*备份用户表空间*/

 

alter tablespace users begin backup;

 

 

 

select * from v$archived_log;

 

 

 

select * from v$controlfile;

 

 

 

alter system setcontrol_files='$ORACLE_HOME/oradata/u01/ctrl01.ctl',

 

'$ORACLE_HOME/oradata/u01/ctrl02.ctl'scope=spfile;

 

 

 

cp $ORACLE_HOME/oradata/u01/ctrl01.ctl$ORACLE_HOME/oradata/u01/ctrl02.ctl

 

 

 

startup pfile='../initSID.ora'

 

 

 

select * from v$parameter where name like'control%' 

 

 

 

show parameter control;

 

 

 

select * from v$controlfile_record_section;

 

 

 

select * from v$tempfile;

 

 

 

/*备份控制文件*/

 

alter database backup controlfile to'../filepath/control.bak';

 

 

 

/*备份控制文件，并将二进制控制文件变为了asc的文本文件*/

 

alter database backup controlfile to trace;

 

 

 

############### redo log ##############

 

 

 

archive log list;

 

 

 

alter system archive log start;--启动自动存档

 

 

 

alter system switch logfile;--强行进行一次日志switch

 

 

 

alter system checkpoint;--强制进行一次checkpoint

 

 

 

alter tablspace users begin backup;

 

 

 

alter tablespace offline;

 

 

 

/*checkpoint 同步频率参数FAST_START_MTTR_TARGET,同步频率越高，系统恢复所需时间越短*/

 

show parameter fast;

 

 

 

show parameter log_checkpoint;

 

 

 

/*加入一个日志组*/

 

alter database add logfile group 3('/$ORACLE_HOME/oracle/ora_log_file6.rdo' size 10M);

 

 

 

/*加入日志组的一个成员*/

 

alter database add logfile member'/$ORACLE_HOME/oracle/ora_log_file6.rdo' to group 3;

 

 

 

/*删除日志组:当前日志组不能删；活动的日志组不能删；非归档的日志组不能删*/

 

alter database drop logfile group 3;

 

 

 

/*删除日志组中的某个成员，但每个组的最后一个成员不能被删除*/

 

alter databse drop logfile member'$ORACLE_HOME/oracle/ora_log_file6.rdo';

 

 

 

/*清除在线日志*/

 

alter database clear logfile'$ORACLE_HOME/oracle/ora_log_file6.rdo';

 

 

 

alter database clear logfile group 3;

 

 

 

/*清除非归档日志*/

 

alter database clear unarchived logfilegroup 3;

 

 

 

/*重命名日志文件*/

 

alter database rename file'$ORACLE_HOME/oracle/ora_log_file6.rdo' to'$ORACLE_HOME/oracle/ora_log_file6a.rdo';

 

 

 

show parameter db_create;

 

 

 

alter system set db_create_online_log_dest_1='path_name';

 

 

 

select * from v$log;

 

 

 

select * from v$logfile;

 

 

 

/*数据库归档模式到非归档模式的互换,要启动到mount状态下才能改变;startupmount;然后再打开数据库.*/

 

alter database noarchivelog/archivelog;

 

 

 

achive log start;---启动自动归档

 

 

 

alter system archive all;－－手工归档所有日志文件

 

 

 

select * from v$archived_log;

 

 

 

show parameter log_archive;

 

 

 

###### 分析日志文件logmnr##############

 

 

 

1) 在init.ora中setutl_file_dir 参数

 

2) 重新启动oracle

 

3) create 目录文件

 

desc dbms_logmnr_d;

 

dbms_logmnr_d.build;

 

4) 加入日志文件 add/remove logfile

 

dhms_logmnr.add_logfile

 

dbms_logmnr.removefile

 

5) start logmnr

 

dbms_logmnr.start_logmnr

 

6) 分析出来的内容查询v$logmnr_content --sqlredo/sqlundo

 

 

 

实践：

 

 

 

desc dbms_logmnr_d;

 

 

 

/*对数据表做一些操作，为恢复操作做准备*/

 

update 表 set qty=10 wherestor_id=6380;

 

 

 

delete 表 where stor_id=7066;

 

/***********************************/

 

utl_file_dir的路径

 

executedbms_logmnr_d.build('foxdict.ora','$ORACLE_HOME/oracle/admin/fox/cdump');

 

 

 

execute dbms_logmnr.add_logfile('$ORACLE_HOME/oracle/ora_log_file6.log',dbms_logmnr.newfile);

 

 

 

executedbms_logmnr.start_logmnr(dictfilename=>'$ORACLE_HOME/oracle/admin/fox/cdump/foxdict.ora');

 

 

 

######### tablespace ##############

 

 

 

select * form v$tablespace;

 

 

 

select * from v$datafile;

 

 

 

/*表空间和数据文件的对应关系*/

 

select t1.name,t2.name from v$tablespacet1,v$datafile t2 where t1.ts#=t2.ts#;

 

 

 

alter tablespace users add datafile 'path'size 10M;

 

 

 

select * from dba_rollback_segs;

 

 

 

/*限制用户在某表空间的使用限额*/

 

alter user user_name quota 10m ontablespace_name;

 

 

 

create tablespace xxx [datafile'path_name/datafile_name'] [size xxx] [extent management local/dictionary][default storage(xxx)];

 

 

 

exmple: create tablespace userdata datafile'$ORACLE_HOME/oradata/userdata01.dbf' size 100M AUTOEXTEND ON NEXT 5M MAXSIZE200M;

 

create tablespace userdata datafile'$ORACLE_HOME/oradata/userdata01.dbf' size 100M extent management dictionarydefault storage(initial 100k next 100k pctincrease 10) offline;

 

/*9i以后，oracle建议使用local管理，而不使用dictionary管理，因为local采用bitmap管理表空间 ，不会产生系统表空间的自愿争用;*/

 

create tablespace userdata datafile'$ORACLE_HOME/oradata/userdata01.dbf' size 100M extent management local uniformsize 1m;

 

create tablespace userdata datafile'$ORACLE_HOME/oradata/userdata01.dbf' size 100M extent management localautoallocate;

 

/*在创建表空间时，设置表空间内的段空间管理模式，这里用的是自动管理*/

 

create tablespace userdata datafile'$ORACLE_HOME/oradata/userdata01.dbf' size 100M extent management local uniformsize 1m segment space management auto;

 

 

 

alter tablespace userdata mininum extent10;

 

 

 

alter tablespace userdata defaultstorage(initial 1m next 1m pctincrease 20);

 

 

 

/*undo tablespace(不能被用在字典管理模下)*/

 

create undo tablespace undo1 datafile'$ORACLE_HOME/oradata/undo101.dbf' size 40M extent management local;

 

 

 

show parameter undo;

 

 

 

/*temporary tablespace*/

 

create temporary tablespace userdatatempfile '$ORACLE_HOME/oradata/undo101.dbf' size 10m extent management local;

 

 

 

/*设置数据库缺省的临时表空间*/

 

alter database default temporary tablespacetablespace_name;

 

 

 

/*系统/临时/在线的undo表空间不能被offline*/

 

alter tablespace tablespace_nameoffline/online;

 

 

 

alter tablespace tablespace_name read only;

 

 

 

/*重命名用户表空间*/

 

alter tablespace tablespace_name renamedatafile '$ORACLE_HOME/oradata/undo101.dbf' to'$ORACLE_HOME/oradata/undo102.dbf';

 

 

 

/*重命名系统表空间 ,但在重命名前必须将数据库shutdown,并重启到mount状态*/

 

alter database rename file'$ORACLE_HOME/oradata/system01.dbf' to '$ORACLE_HOME/oradata/system02.dbf';

 

 

 

drop tablespace userdata including contentsand datafiles;---drop tablespce

 

 

 

/*resize tablespace,autoextend datafilespace*/

 

alter database datafile'$ORACLE_HOME/oradata/undo102.dbf' autoextend on next 10m maxsize 500M;

 

 

 

/*resize datafile*/

 

alter database datafile'$ORACLE_HOME/oradata/undo102.dbf' resize 50m;

 

 

 

/*给表空间扩展空间*/

 

alter tablespace userdata add datafile'$ORACLE_HOME/oradata/undo102.dbf' size 10m;

 

 

 

/*将表空间设置成OMF状态*/

 

alter system setdb_create_file_dest='$ORACLE_HOME/oradata';

 

 

 

create tablespace userdata;---use OMFstatus to create tablespace;

 

 

 

drop tablespace userdata;---user OMF statusto drop tablespace;

 

 

 

select * fromdba_tablespace/v$tablespace/dba_data_files;

 

 

 

/*将表的某分区移动到另一个表空间*/

 

alter table table_name move partitionpartition_name tablespace tablespace_name;

 

 

 

###### ORACLE storage structure andrelationships #########

 

 

 

/*手工分配表空间段的分区(extend)大小*/

 

alter table kong.test12 allocateextent(size 1m datafile '$ORACLE_HOME/oradata/undo102.dbf');

 

 

 

alter table kong.test12 deallocate unused;---释放表中没有用到的分区

 

 

 

show parameter db;

 

 

 

alter system set db_8k_cache_size=10m; ---配置8k块的内存空间块参数

 

 

 

select * fromdba_extents/dba_segments/data_tablespace;

 

 

 

select * fromdba_free_space/dba_data_file/data_tablespace;

 

 

 

/*数据对象所占用的字节数*/

 

select sum(bytes) from dba_extents whereonwer='kong' and segment_name ='table_name';

 

 

 

############ UNDO Data ################

 

 

 

show parameter undo;

 

 

 

alter tablespace users offline normal;

 

 

 

alter tablespace users offline immediate;

 

 

 

recover datafile'$ORACLE_HOME/oradata/undo102.dbf';

 

 

 

alter tablespace users online 

 

 

 

select * from dba_rollback_segs;

 

 

 

alter system set undo_tablespace=undotbs1;

 

 

 

/*忽略回滚段的错误提示*/

 

alter system set undo_suppress_errors=true;

 

 

 

/*在自动管理模式下,不会真正建立rbs1;在手工管理模式则可以建立,且是私有回滚段*/

 

create rollback segment rbs1 tablespaceundotbs;

 

 

 

desc dbms_flashback;

 

 

 

/*在提交了修改的数据后,9i提供了旧数据的回闪操作,将修改前的数据只读给用户看,但这部分数据不会又恢复在表中,而是旧数据的一个映射*/

 

executedbms_flashback.enable_at_time('26-JAN-04:12:17:00 pm');

 

 

 

execute dbms_flashback.disable;

 

 

 

/*回滚段的统计信息*/

 

select end_time,begin_time,undoblks fromv$undostat;

 

 

 

/*undo表空间的大小计算公式:UndoSpace=[UR * (UPS * DBS)] + (DBS * 24)

 

UR :UNDO_RETENTION 保留的时间(秒)

 

UPS :每秒的回滚数据块

 

DBS:系统EXTENT和FILE SIZE(也就是db_block_size)*/

 

 

 

select * fromdba_rollback_segs/v$rollname/v$rollstat/v$undostat/v$session/v$transaction;

 

 

 

show parameter transactions;

 

 

 

show parameter rollback;

 

 

 

/*在手工管理模式下,建立公共的回滚段*/

 

create public rollback segment prbs1tablespace undotbs;

 

 

 

alter rollback segment rbs1 online;----在手工管理模式

 

 

 

/*在手工管理模式中,initSID.ora中指定undo_management=manual 、rollback_segment=('rbs1','rbs2',...)、

 

transactions=100 、transactions_per_rollback_segment=10

 

然后 shutdown immediate ,startup pfile=....\???.ora */

 

 

 

########## Managing Tables ###########

 

 

 

/*char type maxlen=2000;varchar2 typemaxlen=4000 bytes

 

rowid 是18位的64进制字符串 (10个bytes 80bits)

 

rowid组成: object#(对象号)--32bits,6位

 

rfile#(相对文件号)--10bits,3位

 

block#(块号)--22bits,6位

 

row#(行号)--16bits,3位

 

64进制: A-Z,a-z,0-9,/,+ 共64个符号

 

 

 

dbms_rowid 包中的函数可以提供对rowid的解释*/

 

 

 

selectrowid,dbms_rowid.rowid_block_number(rowid),dbms_rowid.rowid_row_number(rowid)from table_name;

 

 

 

create table test2

 

(

 

id int,

 

lname varchar2(20) not null,

 

fname varchar2(20) constraint ck_1check(fname like 'k%'),

 

empdate date default sysdate)

 

) tablespace tablespace_name;

 

 

 

 

 

create global temporary table test2 oncommit delete/preserve rows as select * from kong.authors;

 

 

 

create table user.table(...) tablespacetablespace_name storage(...) pctfree10 pctused 40;

 

 

 

alter table user.tablename pctfree 20pctused 50 storage(...);---changing table storage

 

 

 

/*手工分配分区,分配的数据文件必须是表所在表空间内的数据文件*/

 

alter table user.table_name allocateextent(size 500k datafile '...');

 

 

 

/*释放表中没有用到的空间*/

 

alter table table_name deallocate unused;

 

 

 

alter table table_name deallocate unusedkeep 8k;

 

 

 

/*将非分区表的表空间搬到新的表空间,在移动表空间后，原表中的索引对象将会不可用，必须重建*/

 

alter table user.table_name move tablespacenew_tablespace_name;

 

 

 

create index index_name onuser.table_name(column_name) tablespace users;

 

 

 

alter index index_name rebuild;

 

 

 

drop table table_name [CASCADECONSTRAINTS];

 

 

 

alter table user.table_name drop columncol_name [CASCADE CONSTRAINTS CHECKPOINT 1000];---drop column

 

 

 

/*给表中不用的列做标记*/

 

alter table user.table_name set unusedcolumn comments CASCADE CONSTRAINTS;

 

 

 

/*drop表中不用的做了标记列*/

 

alter table user.table_name drop unusedcolumns checkpoint 1000;

 

 

 

/*当在drop col是出现异常，使用CONTINUE，防止重删前面的column*/

 

ALTER TABLE USER.TABLE_NAME DROP COLUMNSCONTINUE CHECKPOINT 1000;

 

 

 

select * from dba_tables/dba_objects;

 

 

 

######## managing indexes ##########

 

 

 

/*create index*/

 

example:

 

/*创建一般索引*/

 

create index index_name ontable_name(column_name) tablespace tablespace_name;

 

/*创建位图索引*/

 

create bitmap index index_name ontable_name(column_name1,column_name2) tablespace tablespace_name;

 

/*索引中不能用pctused*/

 

create [bitmap] index index_name ontable_name(column_name) tablespace tablespace_name pctfree 20 storage(inital100k next 100k) 

 

/*大数据量的索引最好不要做日志*/

 

create [bitmap] index index_nametable_name(column_name1,column_name2) tablespace_name pctfree 20 storage(inital100k next 100k) nologging;

 

/*创建反转索引*/

 

create index index_name ontable_name(column_name) reverse;

 

/*创建函数索引*/

 

create index index_name ontable_name(function_name(column_name)) tablespace tablespace_name;

 

/*建表时创建约束条件*/

 

create table user.table_name(column_namenumber(7) constraint constraint_name primary key deferrable using indexstorage(initial 100k next 100k) tablespace tablespace_name,column_name2varchar2(25) constraint constraint_name not null,column_name3 number(7))tablespace tablespace_name;

 

 

 

/*给创建bitmap index分配的内存空间参数，以加速建索引*/

 

show parameter create_bit;

 

 

 

/*改变索引的存储参数*/

 

alter index index_name pctfree 30storage(initial 200k next 200k);

 

 

 

/*给索引手工分配一个分区*/

 

alter index index_name allocate extent(size 200k datafile '$ORACLE/oradata/..');

 

 

 

/*释放索引中没用的空间*/

 

alter index index_name deallocate unused;

 

 

 

/*索引重建*/

 

alter index index_name rebuild tablespacetablespace_name;

 

 

 

/*普通索引和反转索引的互换*/

 

alter index index_name rebuild tablespacetablespace_name reverse;

 

 

 

/*重建索引时，不锁表*/

 

alter index index_name rebuild online;

 

 

 

/*给索引整理碎片*/

 

alter index index_name COALESCE;

 

 

 

/*分析索引,事实上是更新统计的过程*/

 

analyze index index_name validatestructure;

 

 

 

desc index_state;

 

 

 

drop index index_name;

 

 

 

alter index index_name monitoringusage;-----监视索引是否被用到

 

 

 

alter index index_name nomonitoringusage;----取消监视

 

 

 

/*有关索引信息的视图*/

 

select * fromdba_indexes/dba_ind_columns/dbs_ind_expressions/v$object_usage;

 

 

 

########## 数据完整性的管理(Maintainingdata integrity) ##########

 

 

 

alter table table_name drop constraintconstraint_name;----drop 约束

 

 

 

alter table table_name add constraintconstraint_name primary key(column_name1,column_name2);-----创建主键

 

 

 

alter table table_name add constraintconstraint_name unique(column_name1,column_name2);---创建唯一约束

 

 

 

/*创建外键约束*/

 

alter table table_name add constraintconstraint_name foreign key(column_name1) references table_name(column_name1);

 

 

 

/*不效验老数据，只约束新的数据[enable/disable：约束/不约束新数据;novalidate/validate:不对/对老数据进行验证]*/

 

alter table table_name add constraintconstraint_name check(column_name like 'B%') enable/disablenovalidate/validate;

 

 

 

/*修改约束条件，延时验证，commit时验证*/

 

alter table table_name modify constraintconstraint_name initially deferred;

 

 

 

/*修改约束条件，立即验证*/

 

alter table table_name modify constraintconstraint_name initially immediate;

 

 

 

alter session setconstraints=deferred/immediate;

 

 

 

/*drop一个有外键的主键表,带cascadeconstraints参数级联删除*/

 

drop table table_name cascade constraints;

 

 

 

/*当truncate外键表时，先将外键设为无效，再truncate;*/

 

truncate table table_name;

 

 

 

/*设约束条件无效*/

 

alter table table_name disable constraintconstraint_name;

 

 

 

alter table table_name enable novalidateconstraint constraint_name;

 

 

 

/*将无效约束的数据行放入exception的表中，此表记录了违反数据约束的行的行号；在此之前，要先建exceptions表*/

 

alter table table_name add constraintconstraint_name check(column_name >15) enable validate exceptions intoexceptions;

 

 

 

/*运行创建exceptions表的脚本*/

 

start $ORACLE_HOME/rdbms/admin/utlexcpt.sql;

 

 

 

/*获取约束条件信息的表或视图*/

 

select * fromuser_constraints/dba_constraints/dba_cons_columns;

 

 

 

################## managing passwordsecurity and resources ####################

 

 

 

alter user user_name accountunlock/open;----锁定/打开用户;

 

 

 

alter user user_name password expire;---设定口令到期

 

 

 

/*建立口令配置文件,failed_login_attempts口令输多少次后锁，password_lock_times指多少天后口令被自动解锁*/

 

create profile profile_name limitfailed_login_attempts 3 password_lock_times 1/1440;

 

/*创建口令配置文件*/

 

create profile profile_name limitfailed_login_attempts 3 password_lock_time unlimited password_life_time 30password_reuse_time 30 password_verify_function verify_functionpassword_grace_time 5;

 

/*建立资源配置文件*/

 

create profile prfile_name limitsession_per_user 2 cpu_per_session 10000 idle_time 60 connect_time 480;

 

 

 

alter user user_name profile profile_name;

 

 

 

/*设置口令解锁时间*/

 

alter profile profile_name limitpassword_lock_time 1/24;

 

 

 

/*password_life_time指口令文件多少时间到期，password_grace_time指在第一次成功登录后到口令到期有多少天时间可改变口令*/

 

alter profile profile_name limitpassword_lift_time 2 password_grace_time 3;

 

 

 

/*password_reuse_time指口令在多少天内可被重用,password_reuse_max口令可被重用的最大次数*/

 

alter profile profile_name limitpassword_reuse_time 10[password_reuse_max 3];

 

 

 

alter user user_name identified byinput_password;-----修改用户口令

 

 

 

drop profile profile_name;

 

 

 

/*建立了profile后，且指定给某个用户，则必须用CASCADE才能删除*/

 

drop profile profile_name CASCADE;

 

 

 

alter system set resource_limit=true;---启用自愿限制,缺省是false

 

 

 

/*配置资源参数*/

 

alter profile profile_name limitcpu_per_session 10000 connect_time 60 idle_time 5;

 

/*资源参数(session级)

 

cpu_per_session 每个session占用cpu的时间 单位1/100秒

 

sessions_per_user 允许每个用户的并行session数

 

connect_time 允许连接的时间 单位分钟

 

idle_time 连接被空闲多少时间后，被自动断开 单位分钟

 

logical_reads_per_session 读块数

 

private_sga 用户能够在SGA中使用的私有的空间数 单位bytes

 

 

 

(call级)

 

cpu_per_call 每次(1/100秒)调用cpu的时间

 

logical_reads_per_call 每次调用能够读的块数

 

*/

 

 

 

alter profile profile_name limitcpu_per_call 1000 logical_reads_per_call 10;

 

 

 

desc dbms_resouce_manager;---资源管理器包

 

 

 

/*获取资源信息的表或视图*/

 

select * from dba_users/dba_profiles;

 

 

 

###### Managing users ############

 

 

 

show parameter os;

 

 

 

create user testuser1 identified bykxf_001;

 

 

 

grant connect,createtable to testuser1;

 

 

 

alter user testuser1 quota 10m ontablespace_name;

 

 

 

/*创建用户*/

 

create user user_name identified bypassword default tablespace tablespace_name temporary tablespacetablespace_name quota 15m on tablespace_name password expire;

 

 

 

/*数据库级设定缺省临时表空间*/

 

alter database default temporary tablespacetablespace_name;

 

 

 

/*制定数据库级的缺省表空间*/

 

alter database default tablespacetablespace_name;

 

 

 

/*创建os级审核的用户，需知道os_authent_prefix，表示oracle和os口令对应的前缀,'OPS$'为此参数的值，此值可以任意设置*/

 

create user user_name identified byexternally default OPS$tablespace_name tablespace_name temporary tablespacetablespace_name quota 15m on tablespace_name password expire;

 

 

 

/*修改用户使用表空间的限额,回滚表空间和临时表空间不允许授予限额*/

 

alter user user_name quota 5m ontablespace_name;

 

 

 

/*删除用户或删除级联用户(用户对象下有对象的要用CASCADE，将其下一些对象一起删除)*/

 

drop user user_name [CASCADE];

 

 

 

/*每个用户在哪些表空间下有些什么限额*/

 

desc dba_ts_quotas;select * fromdba_ts_quotas where username='...';

 

 

 

/*改变用户的缺省表空间*/

 

alter user user_name default tablespacetablespace_name;

 

 

 

######### Managing Privileges #############

 

 

 

grant create table,create session touser_name;

 

 

 

grant create any table to user_name; revokecreate any table from user_name;

 

 

 

/*授予权限语法,public 标识所有用户,withadmin option允许能将权限授予第三者的权限*/

 

grant system_privs,[......] to[user/role/public],[....] [with admin option];

 

 

 

select * from v$pwfile_users;

 

 

 

/*当O7_dictionary_accessiblity参数为True时，标识select any table时，包括系统表也能select ,否则，不包含系统表;缺省为false*/

 

show parameter O7;

 

 

 

/*由于O7_dictionary_accessiblity为静态参数，不能动态改变，故加scope=spfile,下次启动时才生效*/

 

alter system setO7_dictionary_accessiblity=true scope=spfile;

 

 

 

/*授予对象中的某些字段的权限，如select某表中的某些字段的权限*/

 

grant [object_privs(column,....)],[...] onobject_name to user/role/public,... with grant option;

 

 

 

/*oracle不允许授予select某列的权限,但可以授insert,update某列的权限*/

 

grant insert(column_name1,column_name2,...)on table_name to user_name with grant option;

 

 

 

select * fromdba_sys_privs/session_privs/dba_tab_privs/user_tab_privs/dba_col_privs/user_col_privs;

 

 

 

/*db/os/none 审计被记录在 数据库/操作系统/不审计 缺省是none*/

 

show parameter audit_trail;

 

 

 

/*启动对表的select动作*/

 

audit select on user.table_name by session;

 

 

 

/*by session在每个session中发出command只记录一次，byaccess则每个command都记录*/

 

audit [create table][select/update/inserton object by session/access][whenever successful/not successful];

 

 

 

desc dbms_fga;---进一步设计，则可使用dbms_fgs包

 

 

 

/*取消审计*/

 

noaudit select on user.table_name;

 

 

 

/*查被审计信息*/

 

select * fromall_def_audit_opts/dba_stmt_audit_opts/dba_priv_audit_opts/dba_obj_audit_opts;

 

 

 

/*获取审计记录*/

 

select * fromdba_audit_trail/dba_audit_exists/dba_audit_object/dba_audit_session/dba_audit_statement;

 

 

 

########### Managing Role #################

 

 

 

create role role_name; grant select ontable_name to role_name; grant role_name to user_name; set role role_name;

 

 

 

create role role_name;

 

create role role_name identified bypassword;

 

create role role_name identifiedexternally;

 

 

 

set role role_name  ----激活role

 

set role role_name identified by password;

 

 

 

alter role role_name not identified;

 

alter role role_name identified bypassword;

 

alter role role_name identified externally;

 

 

 

grant priv_name to role_name [WITH ADMINOPTION];

 

grant update(column_name1,col_name2,...) ontable_name to role_name;

 

grant role_name1 to role_name2;

 

 

 

/*建立default role,用户登录时，缺省激活defaultrole*/

 

alter user user_name default rolerole_name1,role_name2,...;

 

alter user user_name default role all;

 

alter user user_name default role allexcept role_name1,...;

 

alter user user_name default role none;

 

 

 

set role role1 [identified bypassword],role2,....;

 

set role all;

 

set role except role1,role2,...;

 

set role none;

 

 

 

revoke role_name from user_name;

 

revoke role_name from public;

 

 

 

drop role role_name;

 

 

 

select * fromdba_roles/dba_role_privs/role_role_privs/dba_sys_privs/role_sys_privs/role_tab_privs/session_roles;

 

 

 

########### Basic SQL SELECT################

 

 

 

select col_name as col_alias fromtable_name 

 

 

 

select col_name from table_name where col1like '_o%'; ----'_'匹配单个字符

 

 

 

/*使用字符函数(右边截取,字段中包含某个字符,左边填充某字符到固定位数,右边填充某字符到固定位数)*/

 

selectsubstr(col1,-3,5),instr(col2,'g'),LPAD(col3,10,'$'),RPAD(col4,10,'%') fromtable_name;

 

 

 

/*使用数字函数(往右/左几位四舍五入,取整,取余)*/

 

select round(col1,-2),trunc(col2),mod(col3)from table_name 

 

 

 

/*使用日期函数(计算两个日期间相差几个星期,两个日期间相隔几个月,在某个月份上加几个月,某个日期的下一个日期,

 

某日期所在月的最后的日期,对某个日期的月分四舍五入，对某个日期的月份进行取整)*/

 

select (sysdate-col1)/7week,months_between(sysdate,col1),add_months(col1,2),next_day(sysdate,'FRIDAY'),last_day(sysdate),

 

round(sysdate,'MONTH'),trunc(sysdate,'MONTH')from table_name;

 

 

 

/*使用NULL函数(当expr1为空取expr2/当expr1为空取expr2,否则取expr3/当expr1=expr2返回空)*/

 

selectnvl(expr1,expr2),nvl2(expr1,expr2,expr3),nullif(expr1,expr2) from table_name;

 

 

 

select column1,column2,column3, casecolumn2 when '50' then column2*1.1

 

when '30' then column2*2.1

 

when '10' then column3/20

 

else column3

 

end as ttt

 

from table_name  ------使用case函数

 

 

 

select table1.col1,table2.col2 from table1

 

[CROSS JOIN table2] | -----笛卡儿连接

 

[NATURAL JOIN table2] | -----用两个表中的同名列连接

 

[JOIN table2 USING (column_name)] | -----用两个表中的同名列中的某一列或几列连接

 

[JOIN table2

 

ON (table1.col1=table2.col2)] |

 

[LEFT|RIGHT|FULL OUTER JOIN table2 ------相当于(+)=,=(+)连接,全外连接

 

ON (table1.col1=table2.col2)]; ------SQL1999中的JOIN语法;

 

 

 

example:

 

select col1,col2 from table1 t1

 

join table2 t2

 

on t1.col1=t2.col2 and t1.col3=t2.col1

 

join table3 t3

 

on t2.col1=t3.col3;

 

 

 

select * from table_name where col1 <any (select col2 from table_name2 where continue group by col3);

 

 

 

select * from table_name where col1 <all (select col2 from table_name2 where continue group by col3);

 

 

 

insert into (select col1,col2,col3 formtable_name where col1> 50 with check option) values (value1,value2,value3);

 

 

 

MERGE INTO table_name table1

 

USING table_name2 table2

 

ON (table1.col1=table2.col2)

 

WHEN MATCHED THEN

 

UPDATE SET

 

table1.col1=table2.col2,

 

table1.col2=table2.col3,

 

...

 

WHEN NOT MATCHED THEN

 

INSERTvalueS(table2.col1,table2.col2,table2.col3,...); -----合并语句

 

 

 

##################### CREATE/ALTER TABLE#######################

 

 

 

alter table table_name drop columncolumn_name ---drop column

 

 

 

alter table table_name set unused(col1,col2,...);----设置列无效，这个比较快。

 

alter table table_name drop unusedcolumns;---删除被设为无效的列

 

 

 

rename table_name1 to table_name2; ---重命名表

 

 

 

comment on table table_name is 'commentmessage';----给表放入注释信息

 

 

 

create table table_name

 

(col1 int not null,col2 varchar2(20),col3varchar2(20),

 

constraint uk_test2_1 unique(col2,col3)));-----定义表中的约束条件

 

 

 

alter table table_name add constraintpk_test2 primary key(col1,col2,...); ----创建主键

 

 

 

/*建立外键*/

 

create table table_name (rid int,namevarchar2(20),constraint fk_test3 foreign key(rid) referencesother_table_name(id));

 

 

 

alter table table_name add constraintck_test3 check(name like 'K%');

 

 

 

alter table table_name drop constraintconstraint_name;

 

 

 

alter table table_name drop primary keycascade;----级联删除主键

 

 

 

alter table table_name disable/enableconstraint constraint_name;----使约束暂时无效

 

 

 

/*删除列，并级联删除此列下的约束条件*/

 

alter table table_name drop columncolumn_name cascade constraint;

 

 

 

select * fromuser_constraints/user_cons_columns;---约束条件相关视图

 

 

 

############## Create Views#####################

 

 

 

CREATE [OR REPLACE] [FORCE|NOFORCE] VIEWview_name [(alias[,alias]...)]

 

AS subquery

 

[WITH CHECK OPTION [CONSTRAINTconstraint_name]]

 

[WITH READ ONLY [CONSTRAINTconstraint_name]]; ------创建视图的语法

 

 

 

example: Create or replace view testview asselect col1,col2,col3 from table_name; ------创建视图

 

/*使用别名*/

 

Create or replace view testview as selectcol1,sum(col2) col2_alias from table_name;

 

/*创建复杂视图*/

 

Create view view_name(alias1,alias2,alias3,alias4) as selectd.col1,min(e.col1),max(e.col1),avg(e.col1) from table_name1 e,table_name2 dwhere e.col2=d.col2 group by d.col1;

 

/*当用update修改数据时，必须满足视图的col1>10的条件，不满足则不能被改变.*/

 

Create or replace view view_name as select* from table_name where col1>10 with check option;

 

 

 

/*改变视图的值.对于简单视图可以用update语法修改表数据，但复杂视图则不一定能改。如使用了函数，group by ,distinct等的列*/

 

update view_name set col1=value1;

 

 

 

/*TOP-N分析*/

 

select [column_list],rownum from (select[column_list] from table_name order by Top-N_column) where rownum<=N;

 

 

 

/*找出某列三条最大值的记录*/

 

example: select rownum as rank ,col1 ,col2from (select col1 ,col2 from table_name order by col2 desc) where rownum<=3;

 

 

 

############# Other database Object###############

 

 

 

CREATE SEQUENCE sequence_name [INCREMENT BYn]

 

[START WITH n]

 

[{MAXvalue n | NOMAXvalue}]

 

[{MINvalue n | NOMINvalue}]

 

[{CYCEL | NOCYCLE}]

 

[{CACHE n | NOCACHE}]; -----创建SEQUENCE

 

 

 

example:

 

CREATE SEQUENCE sequence_name INCREMENT BY10

 

START WITH 120

 

MAXvalue 9999

 

NOCACHE

 

NOCYCLE;

 

 

 

select * from user_sequences ---当前用户下记录sequence的视图

 

 

 

selectsequence_name.nextval,sequence_name.currval from dual;-----sequence的引用

 

 

 

alter sequence sequence_name INCREMENT BY20

 

MAXvalue 999999

 

NOCACHE

 

NOCYCLE; -----修改sequence,不能改变起始序号

 

 

 

drop sequence sequence_name; ----删除sequence

 

 

 

CREATE [PUBLIC] SYNONYM synonym_name FORobject; ------创建同义词

 

 

 

DROP [PUBLIC] SYNONYM synonym_name;----删除同义词

 

 

 

CREATE PUBLIC DATABASE LINK link_nameUSEING OBJECT;----创建DBLINK

 

 

 

select * from object_name@link_name; ----访问远程数据库中的对象

 

 

 

/*union 操作，它将两个集合的交集部分压缩，并对数据排序*/

 

select col1,col2,col3 from table1_nameunion select col1,col2,col3 from table2_name;

 

 

 

/*union all 操作，两个集合的交集部分不压缩，且不对数据排序*/

 

select col1,col2,col3 from table1_nameunion all select col1,col2,col3 from table2_name;

 

 

 

/*intersect 操作，求两个集合的交集,它将对重复数据进行压缩，且排序*/

 

select col1,col2,col3 from table1_nameintersect select col1,col2,col3 from table2_name;

 

 

 

/*minus 操作，集合减,它将压缩两个集合减后的重复记录, 且对数据排序*/

 

select col1,col2,col3 from table1_nameminus select col1,col2,col3 from table2_name;

 

 

 

/*EXTRACT 抽取时间函数. 此例是抽取当前日期中的年*/

 

select EXTRACT(YEAR FROM SYSDATE) fromdual;

 

/*EXTRACT 抽取时间函数. 此例是抽取当前日期中的月*/

 

select EXTRACT(MONTH FROM SYSDATE) fromdual;

 

 

 

########################## 增强的 group by 子句#########################

 

 

 

select [column,] group_function(column)...

 

from table

 

[WHERE condition]

 

[GROUP BY [ROLLUP] group_by_expression]

 

[HAVING having_expression];

 

[ORDER BY column]; -------ROLLUP操作字，对group by子句的各字段从右到左进行再聚合

 

 

 

example:

 

/*其结果看起来象对col1做小计*/

 

select col1,col2,sum(col3) from table groupby rollup(col1,col2);

 

/*复合rollup表达式*/

 

select col1,col2,sum(col3) from table groupby rollup((col1,col2));

 

 

 

select [column,] group_function(column)...

 

from table

 

[WHERE condition]

 

[GROUP BY [CUBE] group_by_expression]

 

[HAVING having_expression];

 

[ORDER BY column]; -------CUBE操作字，除完成ROLLUP的功能外，再对ROLLUP后的结果集从右到左再聚合

 

 

 

example:

 

/*其结果看起来象对col1做小计后，再对col2做小计，最后算总计*/

 

select col1,col2,sum(col3) from table groupby cube(col1,col2);

 

/*复合rollup表达式*/

 

select col1,col2,sum(col3) from table groupby cube((col1,col2));

 

/*混合rollup,cube表达式*/

 

select col1,col2,col3,sum(col4) from tablegroup by col1,rollup(col2),cube(col3);

 

 

 

/*GROUPING(expr)函数，查看select语句种以何字段聚合，其取值为0或1*/

 

select [column,]group_function(column)...,GROUPING(expr)

 

from table

 

[WHERE condition]

 

[GROUP BY [ROLLUP] group_by_expression]

 

[HAVING having_expression];

 

[ORDER BY column];

 

 

 

example:

 

selectcol1,col2,sum(col3),grouping(col1),grouping(col2) from table group bycube(col1,col2);

 

 

 

/*grouping sets操作，对group by结果集先对col1求和，再对col2求和，最后将其结果集并在一起*/

 

select col1,col2,sum(col3) from table groupby grouping sets((col1),(col2));