Oracle 12C的第一次异常恢复—文件头坏块

接到第一个使用Oracle 12C作为生产库的恢复救援.有两个业务数据文件报文件头损坏,其他数据文件全部是9月份的一次备份,在当前的条件下,希望我们能够帮他们恢复出来业务文件中的数据
数据库版本信息

SQL> select * from v$version;

BANNER                                                                               CON_ID
-------------------------------------------------------------------------------- ----------
Oracle Database 12c Enterprise Edition Release 12.1.0.1.0 - 64bit Production              0
PL/SQL Release 12.1.0.1.0 - Production                                                    0
CORE    12.1.0.1.0      Production                                                        0
TNS for Linux: Version 12.1.0.1.0 - Production                                            0
NLSRTL Version 12.1.0.1.0 - Production                                                    0

数据库故障
具体脚本请参考:数据库恢复检查脚本(Oracle Database Recovery Check)
控制文件信息
1
控制文件中关于数据文件信息
2
数据文件头信息
3
alert日志报错

Reading datafile '/app/oracle/oradata/freetouch/sales.dbf' for corruption at rdba: 0x00000001 (file 4, block 1)
Reread (file 4, block 1) found same corrupt data (no logical check)
Hex dump of (file 5, block 1) in trace file /app/oracle/diag/rdbms/valuenet/valuenet/trace/valuenet_ora_12384.trc

Corrupt block relative dba: 0x00000001 (file 5, block 1)
Fractured block found during kcvxfh v8
Data in bad block:
 type: 0 format: 2 rdba: 0x00000001
 last change scn: 0x0000.00000000 seq: 0x1 flg: 0x05
 spare1: 0x0 spare2: 0x0 spare3: 0x0
 consistency value in tail: 0x00000001
 check value in block header: 0xa701
 computed block checksum: 0x0

Reading datafile '/app/oracle/oradata/freetouch/drp_200200' for corruption at rdba: 0x00000001 (file 5, block 1)
Reread (file 5, block 1) found same corrupt data (no logical check)
Hex dump of (file 4, block 1) in trace file /app/oracle/diag/rdbms/valuenet/valuenet/trace/valuenet_ora_12384.trc

Corrupt block relative dba: 0x00000001 (file 4, block 1)
Fractured block found during kcvxfh v8
Data in bad block:
 type: 0 format: 2 rdba: 0x00000001
 last change scn: 0x0000.00000000 seq: 0x1 flg: 0x05
 spare1: 0x0 spare2: 0x0 spare3: 0x0
 consistency value in tail: 0x00000001
 check value in block header: 0xa701
 computed block checksum: 0x0

odu无法识别异常文件

[oracle@db odu]$ ./odu

Oracle Data Unloader trial version 4.1.3

Copyright (c) 2008,2009,2010,2011 XiongJun. All rights reserved.


loading default config.......

byte_order little
block_size  8192
db_timezone -7
Invalid db timezone:-7
client_timezone 8
Invalid client timezone:8
asmfile_extract_path /home/oracle/hongye/odu/data
data_path  /home/oracle/hongye/odu/data
lob_path  /home/oracle/hongye/odu/data
charset_name ZHS16GBK
ncharset_name AL16UTF16
output_form  dmp
error at line 10.
lob_storage infile
clob_byte_order big
trace_level 1
delimiter |
unload_deleted yes
file_header_offset 0
is_tru64 no
record_row_addr no
convert_clob_charset yes

load config file 'config.txt' successful
loading default asm disk file ......

can not open file 'asmdisk.txt', error message:No such file or directory.
loading default control file ......

unknown file format '/app/oracle/oradata/freetouch/sales.dbf'
unknown file format '/app/oracle/oradata/freetouch/drp_200200'

 ts#   fn  rfn bsize   blocks bf offset filename
---- ---- ---- ----- -------- -- ------ --------------------------------------------
   1    1    1  8192   194560 N       0 /app/oracle/oradata/freetouch/system01.dbf
   6    2   10  8192    45840 N       0 /app/oracle/oradata/freetouch/example01.dbf
   1    3    3  8192   907520 N       0 /app/oracle/oradata/freetouch/sysaux01.dbf
   4 1024   10  8192        0 N       0 /app/oracle/oradata/freetouch/sales.dbf
   5 1024    9  8192        0 N       0 /app/oracle/oradata/freetouch/drp_200200
   4    6    6  8192   128320 N       0 /app/oracle/oradata/freetouch/users01.dbf
   7    7    7  8192   780288 N       0 /app/oracle/oradata/freetouch/undotbs03.dbf
  11    8    8  8192    25600 N       0 /app/oracle/oradata/freetouch/indx01.dbf
load control file 'control.txt' successful
loading dictionary data......done

loading scanned data......done

dul无法识别异常文件

[oracle@db dul]$ ./dul

Data UnLoader: 10.2.0.5.32 - Internal Only - on Sun Nov  2 23:34:42 2014
with 64-bit io functions

Copyright (c) 1994 2014 Bernard van Duijnen All rights reserved.

 Strictly Oracle Internal Use Only


DUL: Warning: ulimit process stack size is only 33554432
Found db_id = 270587870
Found db_name = VALUENET

DUL: Warning: Cannot verify file number for /app/oracle/oradata/freetouch/sales.dbf
DUL: Warning: First four bytes(76 162 0 0) of block 2 are not the start of a proper data block header
DUL: Warning: Block corruption or configuration error
DUL: Warning: Check db_block_size and/or osd_file_leader_size and/or file offset
DUL: Error: File Number can only be zero for Single Tablespace Datafiles

DUL: Warning: Cannot verify file number for /app/oracle/oradata/freetouch/drp_200200
DUL: Warning: First four bytes(76 162 0 0) of block 2 are not the start of a proper data block header
DUL: Warning: Block corruption or configuration error
DUL: Warning: Check db_block_size and/or osd_file_leader_size and/or file offset
DUL: Error: File Number can only be zero for Single Tablespace Datafiles
DUL> show datafiles;
ts# rf# start   blocks offs open  err file name
  0   1     0   194561    0    1    0 /app/oracle/oradata/freetouch/system01.dbf
  1   3     0   907521    0    1    0 /app/oracle/oradata/freetouch/sysaux01.dbf
  4   6     0   128321    0    1    0 /app/oracle/oradata/freetouch/users01.dbf
  7   7     0   780289    0    1    0 /app/oracle/oradata/freetouch/undotbs03.dbf
 11   8     0    25601    0    1    0 /app/oracle/oradata/freetouch/indx01.dbf
  6  10     0    45841    0    1    0 /app/oracle/oradata/freetouch/example01.dbf

该异常文件使用dul/odu均无法正常识别.证明文件头确实已经损坏

dbv 检测

[oracle@db trace]$ dbv file=/app/oracle/oradata/freetouch/drp_200200

DBVERIFY: Release 12.1.0.1.0 - Production on Sun Nov 2 14:08:34 2014

Copyright (c) 1982, 2013, Oracle and/or its affiliates.  All rights reserved.

DBVERIFY - Verification starting : FILE = /app/oracle/oradata/freetouch/drp_200200


DBVERIFY - Verification complete

Total Pages Examined         : 194560
Total Pages Processed (Data) : 114596
Total Pages Failing   (Data) : 0
Total Pages Processed (Index): 26198
Total Pages Failing   (Index): 0
Total Pages Processed (Other): 37787
Total Pages Processed (Seg)  : 1
Total Pages Failing   (Seg)  : 0
Total Pages Empty            : 15979
Total Pages Marked Corrupt   : 0
Total Pages Influx           : 0
Total Pages Encrypted        : 0
Highest block SCN            : 129603862 (0.129603862)

[oracle@db ~]$ dbv file=/app/oracle/oradata/freetouch/sales.dbf

DBVERIFY: Release 12.1.0.1.0 - Production on Sun Nov 2 23:12:05 2014

Copyright (c) 1982, 2013, Oracle and/or its affiliates.  All rights reserved.

DBVERIFY - Verification starting : FILE = /app/oracle/oradata/freetouch/sales.dbf


DBVERIFY - Verification complete

Total Pages Examined         : 655360
Total Pages Processed (Data) : 294938
Total Pages Failing   (Data) : 0
Total Pages Processed (Index): 233404
Total Pages Failing   (Index): 0
Total Pages Processed (Lob)  : 38
Total Pages Failing   (Lob)  : 0
Total Pages Processed (Other): 23252
Total Pages Processed (Seg)  : 0
Total Pages Failing   (Seg)  : 0
Total Pages Empty            : 103728
Total Pages Marked Corrupt   : 0
Total Pages Influx           : 0
Total Pages Encrypted        : 0
Highest block SCN            : 134665298 (0.134665298)

dbv检测结果无坏块,但是v$datafile_header和alert日志中报坏块,初步判断是由于该文件是bigfile,dbv未检测到文件头坏块,实际该该数据文件头损坏,其他block正常.所幸的是该库有9月份的rman备份(中间归档丢失),因此使用rman还原出来9月份的数据文件,然后使用dd拷贝两个 block(block 0和block 1)到异常文件.

[root@db freetouch]# dd if=/app1/oracle/oradata/freetouch/sales.dbf bs=8192 count=2 of=/tmp/odu/sales.2
2+0 records in
2+0 records out
16384 bytes (16 kB) copied, 0.0176368 s, 929 kB/s
[root@db freetouch]# dd if=/tmp/odu/sales.2 of=/app/oracle/oradata/freetouch/sales.dbf bs=8192 count=2 conv=notrunc 
2+0 records in
2+0 records out
16384 bytes (16 kB) copied, 6.4281e-05 s, 255 MB/s
[root@db freetouch]# dd if=/app1/oracle/oradata/freetouch/drp_200200 bs=8192 count=2 of=/tmp/odu/drp_200200.2
2+0 records in
2+0 records out
16384 bytes (16 kB) copied, 0.0185934 s, 881 kB/s
[root@db freetouch]# dd if=/tmp/odu/drp_200200.2 of=/app/oracle/oradata/freetouch/drp_200200 bs=8192 count=2 conv=notrunc 
2+0 records in
2+0 records out
16384 bytes (16 kB) copied, 6.4419e-05 s, 254 MB/s

尝试恢复数据库

SQL> recover database;
ORA-00283: recovery session canceled due to errors
ORA-01610: recovery using the BACKUP CONTROLFILE option must be done


SQL> recover database using backup controlfile;
ORA-00283: recovery session canceled due to errors
ORA-19909: datafile 4 belongs to an orphan incarnation
ORA-01110: data file 4: '/app/oracle/oradata/freetouch/sales.dbf'

使用bbed修改相关文件头,然后继续恢复
具体见:bbed解决ORA-01190类似方法处理

SQL> recover database using backup controlfile;
ORA-00279: change 129603904 generated at 11/02/2014 19:19:54 needed for thread
1
ORA-00289: suggestion :
/app/oracle/recovery_area/VALUENET/archivelog/2014_11_02/o1_mf_1_1_%u_.arc
ORA-00280: change 129603904 for thread 1 is in sequence #1


Specify log: {<RET>=suggested | filename | AUTO | CANCEL}
cancel;
ORA-00308: cannot open archived log 'cancel;'
ORA-27037: unable to obtain file status
Linux-x86_64 Error: 2: No such file or directory
Additional information: 3


Specify log: {<RET>=suggested | filename | AUTO | CANCEL}
cancel
Media recovery cancelled.
SQL> alter database open resetlogs;
alter database open resetlogs
*
ERROR at line 1:
ORA-01092: ORACLE instance terminated. Disconnection forced
ORA-00600: internal error code, arguments: [2662], [0], [129603911], [0],
[129603913], [29360256], [], [], [], [], [], []
Process ID: 19881
Session ID: 1 Serial number: 3

出现ORA-600[2662]错误,因为scn相差比较小,重启数据库机器,出现ORA-600[4194]错误

SQL> startup pfile='/tmp/pfile.txt' mount
ORACLE instance started.

Total System Global Area 2505338880 bytes
Fixed Size                  2291472 bytes
Variable Size             973080816 bytes
Database Buffers         1526726656 bytes
Redo Buffers                3239936 bytes
Database mounted.
SQL> alter database open;
alter database open
*
ERROR at line 1:
ORA-01113: file 1 needs media recovery
ORA-01110: data file 1: '/app/oracle/oradata/freetouch/system01.dbf'


SQL> recover database ;
ORA-00283: recovery session canceled due to errors
ORA-16433: The database or pluggable database must be opened in read/write
mode.

重建控制文件后继续恢复

SQL> alter database open resetlogs;    
alter database open resetlogs
*
ERROR at line 1:
ORA-01092: ORACLE instance terminated. Disconnection forced
ORA-00600: internal error code, arguments: [4194], [46], [19], [], [], [], [],
[], [], [], [], []
Process ID: 20351
Session ID: 1 Serial number: 3

设置undo_management=MANUAL然后继续恢复

[oracle@db tmp]$ sqlplus / as sysdba

SQL*Plus: Release 12.1.0.1.0 Production on Sun Nov 2 19:29:45 2014

Copyright (c) 1982, 2013, Oracle.  All rights reserved.

Connected to an idle instance.

SQL> startup pfile='/tmp/pfile.txt'
ORACLE instance started.

Total System Global Area 2505338880 bytes
Fixed Size                  2291472 bytes
Variable Size             973080816 bytes
Database Buffers         1526726656 bytes
Redo Buffers                3239936 bytes
Database mounted.
Database opened.

这次的恢复也证明Oracle 12C确实有着越来越多的用户在使用.

发表在 非常规恢复 | 标签为 , , , , , , | 评论关闭

ORA-27069: skgfdisp: 尝试在文件范围外执行 I/O

接到网友技术支持请求,win 2003 ntfs格式文件系统,Oracle 8.1.7版本,主机重启后,数据库无法正常启动,offline datafile 15,数据库open成功,但是datafile 无法正常online,报错为:ORA-27069: skgfdisp: attempt to do I/O beyond the range of the file,请求协助处理

SQL> recover datafile 'D:\ORACLE\ORADATA\ORCL\ZSF_DATA.DBF';
ORA-00283: 恢复会话因错误而取消
ORA-01115: 从文件 15 读取块时出现 IO 错误 (块 # 1030071)
ORA-01110: 数据文件 15: 'D:\ORACLE\ORADATA\ORCL\ZSF_DATA.DBF'
ORA-27069: skgfdisp: 尝试在文件范围外执行 I/O
OSD-04026: 无效的参数经过. (OS 1030071)

使用bbed,成功online datafile 15

Tue Oct 28 16:30:35 2014
ALTER DATABASE RECOVER  datafile 15  
Tue Oct 28 16:30:35 2014
Media Recovery Datafile: 15
Media Recovery Start
Media Recovery Log 
Recovery of Online Redo Log: Thread 1 Group 1 Seq 245110 Reading mem 0
  Mem# 0 errs 0: D:\ORACLE\ORADATA\ORCL\REDO03.LOG
Media Recovery failed with error 1115
ORA-283 signalled during: ALTER DATABASE RECOVER  datafile 15  ...
Tue Oct 28 16:32:53 2014
Shutting down instance (abort)
License high water mark = 6
Instance terminated by USER, pid = 1548
Starting up ORACLE RDBMS Version: 8.1.7.0.0.
System parameters with non-default values:
  processes                = 600
  shared_pool_size         = 52428800
  large_pool_size          = 20971520
  java_pool_size           = 20971520
  control_files            = D:\oracle\oradata\ORCL\control01.ctl, D:\oracle\oradata\ORCL\control02.ctl
  db_block_buffers         = 19200
  db_block_size            = 8192
  compatible               = 8.1.0
  log_buffer               = 32768
  log_checkpoint_interval  = 10000
  log_checkpoint_timeout   = 1800
  db_files                 = 1024
  db_file_multiblock_read_count= 8
  max_enabled_roles        = 30
  remote_login_passwordfile= EXCLUSIVE
  global_names             = TRUE
  distributed_transactions = 500
  instance_name            = ORCL
  service_names            = ORCL
  mts_dispatchers          = (PROTOCOL=TCP)(PRE=oracle.aurora.server.SGiopServer)
  open_links               = 4
  sort_area_size           = 65536
  sort_area_retained_size  = 65536
  db_name                  = ORCL
  open_cursors             = 500
  ifile                    = D:\oracle\admin\ORCL\pfile\init.ora
  os_authent_prefix        = 
  job_queue_processes      = 4
  job_queue_interval       = 10
  parallel_max_servers     = 5
  background_dump_dest     = D:\oracle\admin\ORCL\bdump
  user_dump_dest           = D:\oracle\admin\ORCL\udump
  max_dump_file_size       = 10240
  oracle_trace_collection_name= 
PMON started with pid=2
DBW0 started with pid=3
LGWR started with pid=4
CKPT started with pid=5
SMON started with pid=6
RECO started with pid=7
SNP0 started with pid=8
SNP1 started with pid=9
SNP2 started with pid=10
SNP3 started with pid=11
Tue Oct 28 16:33:01 2014
starting up 1 shared server(s) ...
starting up 1 dispatcher(s) for network address '(ADDRESS=(PARTIAL=YES)(PROTOCOL=TCP))'...
Tue Oct 28 16:33:02 2014
ALTER DATABASE   MOUNT
Tue Oct 28 16:33:06 2014
Successful mount of redo thread 1, with mount id 1389958722.
Tue Oct 28 16:33:06 2014
Database mounted in Exclusive Mode.
Completed: ALTER DATABASE   MOUNT
Tue Oct 28 16:33:49 2014
ALTER DATABASE RECOVER  database until cancel  
Tue Oct 28 16:33:49 2014
Media Recovery Start
Media Recovery Log 
kcrrga: Warning.  Log sequence in archive filename wrapped
to fix length as indicated by %S in LOG_ARCHIVE_FORMAT.
Old log archive with same name might be overwritten.
ORA-279 signalled during: ALTER DATABASE RECOVER  database until cancel  ...
Tue Oct 28 16:34:03 2014
ALTER DATABASE RECOVER    LOGFILE 'D:\ORACLE\ORADATA\ORCL\REDO02.LOG'  
Tue Oct 28 16:34:03 2014
Media Recovery Log D:\ORACLE\ORADATA\ORCL\REDO02.LOG
Incomplete recovery applied all redo ever generated.
Recovery completed through change %s139866389
Media Recovery Complete
Completed: ALTER DATABASE RECOVER    LOGFILE 'D:\ORACLE\ORADA
Tue Oct 28 16:34:29 2014
alter database datafile 15 online

Tue Oct 28 16:34:29 2014
Completed: alter database datafile 15 online
Tue Oct 28 16:34:36 2014
alter database open resetlogs

RESETLOGS is being done without consistancy checks. This may result
in a corrupted database. The database should be recreated.
RESETLOGS after incomplete recovery UNTIL CHANGE 139866389
Tue Oct 28 16:34:38 2014
Thread 1 opened at log sequence 1
  Current log# 2 seq# 1 mem# 0: D:\ORACLE\ORADATA\ORCL\REDO02.LOG
Successful open of redo thread 1.
Tue Oct 28 16:34:38 2014
SMON: enabling cache recovery
Tue Oct 28 16:34:38 2014
Dictionary check beginning
Dictionary check complete
Tue Oct 28 16:34:39 2014
SMON: enabling tx recovery
Tue Oct 28 16:34:44 2014
Completed: alter database open resetlogs

数据库datafile 15 online成功后,客户操作业务继续发生ORA-600[ktsxs_add2]错误

Tue Oct 28 17:07:42 2014
Errors in file D:\oracle\admin\ORCL\udump\ORA02340.TRC:
ORA-00600: 内部错误代码,自变量: [ktsxs_add2], [14], [15], [42534], [5732], [5733], [], []

Tue Oct 28 17:07:53 2014
Errors in file D:\oracle\admin\ORCL\udump\ORA02340.TRC:
ORA-00600: 内部错误代码,自变量: [ktsxs_add2], [14], [15], [42534], [5732], [5733], [], []

Tue Oct 28 17:08:03 2014
Errors in file D:\oracle\admin\ORCL\udump\ORA02340.TRC:
ORA-00600: 内部错误代码,自变量: [ktsxs_add2], [14], [15], [42534], [5732], [5733], [], []

Tue Oct 28 17:08:16 2014
Errors in file D:\oracle\admin\ORCL\udump\ORA02340.TRC:
ORA-00600: 内部错误代码,自变量: [ktsxs_add2], [14], [15], [42534], [5732], [5733], [], []

Tue Oct 28 17:08:23 2014
Errors in file D:\oracle\admin\ORCL\udump\ORA02308.TRC:
ORA-00600: 内部错误代码,自变量: [ktsxs_add2], [14], [15], [42534], [5732], [5733], [], []

Tue Oct 28 17:08:31 2014
Errors in file D:\oracle\admin\ORCL\udump\ORA02340.TRC:
ORA-00600: 内部错误代码,自变量: [ktsxs_add2], [14], [15], [42534], [5732], [5733], [], []

Tue Oct 28 17:08:38 2014
Errors in file D:\oracle\admin\ORCL\udump\ORA02308.TRC:
ORA-00600: 内部错误代码,自变量: [ktsxs_add2], [14], [15], [42534], [5732], [5733], [], []

通过分析相关日志发现是insert插入表报错,很好理解,该库的datafile 15已经超过了系统的限制,现在继续插入数据,因此报错,查询可能异常对象

SQL> col segment_name for a20
SQL> SELECT distinct OWNER, SEGMENT_NAME, SEGMENT_TYPE, A.PARTITION_NAME
  2    FROM DBA_EXTENTS A
  3   WHERE FILE_ID = 15
  4     AND 1030071 <= BLOCK_ID;

OWNER                          SEGMENT_NAME         SEGMENT_TYPE
------------------------------ -------------------- ------------------
PARTITION_NAME
------------------------------
ZSF                            DETAIL               TABLE


ZSF                            DETAIL1              INDEX


ZSF                            DETAIL2              INDEX



OWNER                          SEGMENT_NAME         SEGMENT_TYPE
------------------------------ -------------------- ------------------
PARTITION_NAME
------------------------------
ZSF                            DETAIL3              INDEX


ZSF                            DETAIL4              INDEX


ZSF                            FK_RECI_ORD          INDEX



OWNER                          SEGMENT_NAME         SEGMENT_TYPE
------------------------------ -------------------- ------------------
PARTITION_NAME
------------------------------
ZSF                            PREPAY1              INDEX


ZSF                            RECEDETAIL1          INDEX

创建新表空间

Create tablespace zsf_new datafile  'D:\ORACLE\ORADATA\ORCL\ZSF_DATA_new01.dbf' size 4096m;
alter tablespace zsf_new add datafile 'D:\ORACLE\ORADATA\ORCL\ZSF_DATA_new02.dbf' 
size 128m autoextend on next 128M maxsize 4096m;

迁移异常对象到新表空间

alter table ZSF.DETAIL move tablespace ZSF_new;
alter index ZSF.DETAIL1 rebuild tablespace ZSF_new;
alter index ZSF.DETAIL2 rebuild tablespace ZSF_new;
alter index ZSF.DETAIL3 rebuild tablespace ZSF_new;
alter index ZSF.DETAIL4 rebuild tablespace ZSF_new;
alter index ZSF.FK_RECI_ORD rebuild tablespace ZSF_new;
alter index ZSF.PREPAY1 rebuild tablespace ZSF_new;
alter index ZSF.RECEDETAIL1 rebuild tablespace ZSF_new;

然后对于datafile 15所在表空间增加新文件,因为已经迁移了异常对象,然后resize datafile 15小于8G,关闭自扩展,至此该数据库恢复完成

发表在 非常规恢复 | 标签为 , , , , , , | 评论关闭

11g中 connect by 语句执行计划改变

从10.2.0.3升级到11.2.0.4的朋友,如果细心会发现,以下sql在11.2.0.4中执行效率变低(该sql主要是获取连接用户获取权限信息)
select privilege#,level from sysauth$ connect by grantee#=prior privilege# and privilege#>0 start with grantee#=:1 and privilege#>0
如果你接触是Oracle版本比较多,而且还比较细心,你可能会进一步发现在11.2.0.2中该条sql是:select /*+ connect_by_filtering */ privilege#, level from sysauth$ connect by grantee#=prior privilege# and privilege#>0 start with grantee#=:1 and privilege#>0 也就是说使用了/*+ connect_by_filtering */提示.我这里通过简单测试说明问题.

在11.2.0.4环境中

14:16:19 SQL> set autot trace exp stat
14:16:20 SQL> set time on
14:16:20 SQL> set timing on
14:16:20 SQL> var a1 number;
14:16:20 SQL> exec :a1:=6;

PL/SQL 过程已成功完成。

已用时间:  00: 00: 00.00
14:16:20 SQL> select privilege#,level from sysauth$ connect by grantee#=prior 
14:16:20 SQL> privilege# and privilege#>0 start with grantee#=:a1  and privilege#>0
14:16:22 SQL> /

已用时间:  00: 00: 00.01

执行计划
----------------------------------------------------------
Plan hash value: 2624122540

------------------------------------------------------------------------------------------------------
| Id  | Operation                               | Name       | Rows  | Bytes | Cost (%CPU)| Time     |
------------------------------------------------------------------------------------------------------
|   0 | SELECT STATEMENT                        |            |     7 |   182 |     3  (34)| 00:00:01 |
|*  1 |  CONNECT BY NO FILTERING WITH START-WITH|            |       |       |            |          |
|   2 |   INDEX FAST FULL SCAN                  | I_SYSAUTH1 |   618 |  4944 |     2   (0)| 00:00:01 |
------------------------------------------------------------------------------------------------------

Predicate Information (identified by operation id):
---------------------------------------------------

   1 - access("GRANTEE#"=PRIOR "PRIVILEGE#")
       filter("PRIVILEGE#">0 AND "GRANTEE#"=TO_NUMBER(:A1) AND "PRIVILEGE#">0)


统计信息
----------------------------------------------------------
          1  recursive calls
          0  db block gets
          7  consistent gets
          0  physical reads
          0  redo size
        599  bytes sent via SQL*Net to client
        520  bytes received via SQL*Net from client
          2  SQL*Net roundtrips to/from client
          2  sorts (memory)
          0  sorts (disk)
          1  rows processed

这里可以看出来这里使用的执行计划使用了CONNECT BY NO FILTERING WITH START-WITH,逻辑读为7.

10.2.0.3环境中

14:32:57 SQL> set lines 150
14:33:00 SQL> set autot trace exp stat
14:33:01 SQL> set time on
14:33:01 SQL> set timing on
14:33:01 SQL> var a1 number;
14:33:01 SQL> exec :a1:=6;

PL/SQL procedure successfully completed.

Elapsed: 00:00:00.00
14:33:01 SQL> select privilege#,level from sysauth$ connect by grantee#=prior 
14:33:01 SQL> privilege# and privilege#>0 start with grantee#=:a1  and privilege#>0  ;

Elapsed: 00:00:00.00

Execution Plan
----------------------------------------------------------
Plan hash value: 2620769641

----------------------------------------------------------------------------------------
| Id  | Operation                 | Name       | Rows  | Bytes | Cost (%CPU)| Time     |
----------------------------------------------------------------------------------------
|   0 | SELECT STATEMENT          |            |     3 |    24 |     2   (0)| 00:00:01 |
|*  1 |  CONNECT BY WITH FILTERING|            |       |       |            |          |
|*  2 |   INDEX RANGE SCAN        | I_SYSAUTH1 |     3 |    24 |     2   (0)| 00:00:01 |
|   3 |   NESTED LOOPS            |            |       |       |            |          |
|   4 |    CONNECT BY PUMP        |            |       |       |            |          |
|*  5 |    INDEX RANGE SCAN       | I_SYSAUTH1 |     3 |    24 |     2   (0)| 00:00:01 |
----------------------------------------------------------------------------------------

Predicate Information (identified by operation id):
---------------------------------------------------

   1 - access("GRANTEE#"=PRIOR "PRIVILEGE#")
       filter("PRIVILEGE#">0)
   2 - access("GRANTEE#"=TO_NUMBER(:A1) AND "PRIVILEGE#">0)
   5 - access("GRANTEE#"=PRIOR "PRIVILEGE#" AND "PRIVILEGE#">0)


Statistics
----------------------------------------------------------
          0  recursive calls
          0  db block gets
          4  consistent gets
          0  physical reads
          0  redo size
        583  bytes sent via SQL*Net to client
        492  bytes received via SQL*Net from client
          2  SQL*Net roundtrips to/from client
          3  sorts (memory)
          0  sorts (disk)
          1  rows processed

这里执行计划使用的为CONNECT BY WITH FILTERING,而且逻辑读为4,对于这个sql来说,使用CONNECT BY WITH FILTERING执行效率更高.

这里可以很明显的看到:connect by查询的执行计划从10g的CONNECT BY WITH FILTERING变为了11g中的CONNECT BY NO FILTERING WITH SW (UNIQUE),从而使得执行计划发生改变。但是Oracle一般有个特性,就是当引入新特性之时,一般都会伴随隐含参数或者event来屏蔽新特性.这里也例外,我们可以通过”_optimizer_connect_by_elim_dups” = false和”_connect_by_use_union_all” = “old_plan_mode”来屏蔽11g中关于connect by执行计划的改变,使得执行计划恢复到10G的CONNECT BY WITH FILTERING方式

14:30:45 SQL> alter session set "_optimizer_connect_by_elim_dups" = false;

会话已更改。

已用时间:  00: 00: 00.00
14:30:46 SQL> alter session set "_connect_by_use_union_all" = "old_plan_mode";

会话已更改。

已用时间:  00: 00: 00.00
14:30:46 SQL> set autot trace exp stat
14:30:46 SQL> set time on
14:30:46 SQL> set timing on
14:30:46 SQL> var a1 number;
14:30:46 SQL> exec :a1:=6;

PL/SQL 过程已成功完成。

已用时间:  00: 00: 00.00
14:30:46 SQL> select privilege#,level from sysauth$ connect by grantee#=prior 
14:30:46 SQL> privilege# and privilege#>0 start with grantee#=:a1  and privilege#>0 ;

已用时间:  00: 00: 00.01

执行计划
----------------------------------------------------------
Plan hash value: 2620769641

----------------------------------------------------------------------------------------
| Id  | Operation                 | Name       | Rows  | Bytes | Cost (%CPU)| Time     |
----------------------------------------------------------------------------------------
|   0 | SELECT STATEMENT          |            |     2 |    16 |     2   (0)| 00:00:01 |
|*  1 |  CONNECT BY WITH FILTERING|            |       |       |            |          |
|*  2 |   INDEX RANGE SCAN        | I_SYSAUTH1 |     2 |    16 |     2   (0)| 00:00:01 |
|   3 |   NESTED LOOPS            |            |       |       |            |          |
|   4 |    CONNECT BY PUMP        |            |       |       |            |          |
|*  5 |    INDEX RANGE SCAN       | I_SYSAUTH1 |     2 |    16 |     2   (0)| 00:00:01 |
----------------------------------------------------------------------------------------

Predicate Information (identified by operation id):
---------------------------------------------------

   1 - access("GRANTEE#"=PRIOR "PRIVILEGE#")
       filter("PRIVILEGE#">0)
   2 - access("GRANTEE#"=TO_NUMBER(:A1) AND "PRIVILEGE#">0)
   5 - access("GRANTEE#"=PRIOR "PRIVILEGE#" AND "PRIVILEGE#">0)


统计信息
----------------------------------------------------------
          1  recursive calls
          0  db block gets
          4  consistent gets
          0  physical reads
          0  redo size
        599  bytes sent via SQL*Net to client
        520  bytes received via SQL*Net from client
          2  SQL*Net roundtrips to/from client
          3  sorts (memory)
          0  sorts (disk)
          1  rows processed

11.2.0.2中也许是考虑到connect by 不够成熟,因此使用了hint /*+ connect_by_filtering */ 来固定执行计划

14:22:09 SQL> select /*+ connect_by_filtering */ privilege#,level from sysauth$ connect by grantee#=prior
14:22:09 SQL>  privilege# and privilege#>0 start with grantee#=:a1  and privilege#>0 
14:22:10 SQL> /

已用时间:  00: 00: 00.00

执行计划
----------------------------------------------------------
Plan hash value: 2620769641

----------------------------------------------------------------------------------------
| Id  | Operation                 | Name       | Rows  | Bytes | Cost (%CPU)| Time     |
----------------------------------------------------------------------------------------
|   0 | SELECT STATEMENT          |            |     7 |   182 |     8  (25)| 00:00:01 |
|*  1 |  CONNECT BY WITH FILTERING|            |       |       |            |          |
|*  2 |   INDEX RANGE SCAN        | I_SYSAUTH1 |     2 |    16 |     2   (0)| 00:00:01 |
|   3 |   NESTED LOOPS            |            |     5 |   105 |     4   (0)| 00:00:01 |
|   4 |    CONNECT BY PUMP        |            |       |       |            |          |
|*  5 |    INDEX RANGE SCAN       | I_SYSAUTH1 |     2 |    16 |     1   (0)| 00:00:01 |
----------------------------------------------------------------------------------------

Predicate Information (identified by operation id):
---------------------------------------------------

   1 - access("GRANTEE#"=PRIOR "PRIVILEGE#")
       filter("PRIVILEGE#">0)
   2 - access("GRANTEE#"=TO_NUMBER(:A1) AND "PRIVILEGE#">0)
   5 - access("GRANTEE#"="connect$_by$_pump$_002"."prior privilege# " AND
              "PRIVILEGE#">0)


统计信息
----------------------------------------------------------
          1  recursive calls
          0  db block gets
          4  consistent gets
          0  physical reads
          0  redo size
        599  bytes sent via SQL*Net to client
        520  bytes received via SQL*Net from client
          2  SQL*Net roundtrips to/from client
          3  sorts (memory)
          0  sorts (disk)
          1  rows processed

CONNECT BY NO FILTERING WITH SW (UNIQUE)和CONNECT BY WITH FILTERING,没有明显的优劣之分,只有在特定的情况下,进行了实际的测试,选择合适自己的sql的执行计划

发表在 Oracle性能优化 | 标签为 | 评论关闭