• No results found

SQL> RECOVER DATABASE USING BACKUP CONTROLFILE UNTIL CANCEL;

In document Oracle DBA (Page 117-163)

Step 8 : Use RMAN to copy the database from ASM to NON-ASM

93. SQL> RECOVER DATABASE USING BACKUP CONTROLFILE UNTIL CANCEL;

94.

95. ORA-00279: change 7937583 generated at 08/14/2006 20:33:55 needed for thread 1

96. ORA-00289: suggestion : +FLASH_RECOVERY_AREA

97. ORA-00280: change 7937583 for thread 1 is in sequence #36 98.

99.

100. Specify log: {<RET>=suggested | filename | AUTO | CANCEL}

101. CANCEL

102. Media recovery cancelled.

103.

104. SQL> ALTER DATABASE OPEN RESETLOGS;

105.

Database altered.

106. From a SQL*Plus session, re-create any tempfiles that are still

currently on the local file system to ASM. This is done by simply dropping the tempfiles from the local file system and re-creating them in ASM. This

example relies on the initialization

parameter db_create_file_dest=+TESTDB_DATA1:

107. SQL> select tablespace_name, file_name, bytes from dba_temp_files;

108.

109. TABLESPACE_NAME FILE_NAME BYTES 110. --- --- --- 111. TEMP /u02/oradata/TESTDB/datafile/o1_mf_temp_2g17lvcq_.tmp

536870912 112.

113. SQL> alter database tempfile

114. 2 '/u02/oradata/TESTDB/datafile/o1_mf_temp_2g17lvcq_.tmp' 115. 3 drop including datafiles;

116.

117. Database altered.

118.

119. SQL> alter tablespace temp add tempfile size 512m 120. 2 autoextend on next 250m maxsize unlimited;

121.

122. Tablespace altered.

123.

124. SQL> select tablespace_name, file_name, bytes from dba_temp_files;

125.

126. TABLESPACE_NAME FILE_NAME BYTES 127. --- --- ---

TEMP +TESTDB_DATA1/testdb/tempfile/temp.261.598485663 536870912

If users are currently accessing the tempfile(s) you are attempting to drop, you may receive the following error:

SQL> alter database tempfile

2 '/u02/oradata/TESTDB/datafile/o1_mf_temp_2g17lvcq_.tmp' 3 drop including datafiles;

ERROR at line 1:

ORA-25152: TEMPFILE cannot be dropped at this time

As for the poor users who were using the tempfile, their transaction will end and will be greeted with the following error message:

SQL> @testTemp.sql

join dba_extents c on (b.segment_name = c.segment_name) *

ERROR at line 4:

ORA-00372: file 601 cannot be modified at this time ORA-01110: data file 601:

'/u02/oradata/TESTDB/datafile/o1_mf_temp_2g17lvcq_.tmp' ORA-00372: file 601 cannot be modified at this time

ORA-01110: data file 601:

'/u02/oradata/TESTDB/datafile/o1_mf_temp_2g17lvcq_.tmp'

If this happens, you should attempt to drop the tempfile again so the operation is successful:

SQL> alter database tempfile

2 '/u02/oradata/TESTDB/datafile/o1_mf_temp_2g17lvcq_.tmp' 3 drop including datafiles;

Database altered.

128. From a SQL*Plus session, re-create any online redo logfiles that are still currently on the local file system to ASM. This is done by simply dropping the logfiles from the local file system and re-creating them in ASM. This example relies on the initialization

parameters db_create_file_dest=+TESTDB_DATA1 and db_recovery_file_dest

=+FLASH_RECOVERY_AREA:

o Determine the current online redo logfiles to move to ASM by examining the file names (and sizes) from V$LOGFILE:

o SQL> select a.group#, a.member, b.bytes

o 2 from v$logfile a, v$log b where a.group# = b.group#;

o

o GROUP# MEMBER BYTES o --- ---

---

o 1 /u02/oradata/TESTDB/onlinelog/o1_mf_1_2g1g61bm_.log 262144000

o 2 /u02/oradata/TESTDB/onlinelog/o1_mf_2_2g1gd4pr_.log 262144000

o 3 /u02/oradata/TESTDB/onlinelog/o1_mf_3_2g1ghs0t_.log 262144000

o 1

/u02/flash_recovery_area/TESTDB/onlinelog/o1_mf_1_2g1g6bq0_.log 262144000

o 2

/u02/flash_recovery_area/TESTDB/onlinelog/o1_mf_2_2g1gdgn1_.log 262144000

o 3

/u02/flash_recovery_area/TESTDB/onlinelog/o1_mf_3_2g1ghz8z_.log 262144000

o

6 rows selected.

o Force a log switch until the last redo log is marked "CURRENT" by issuing the following command:

o SQL> select group#, status from v$log;

o

o GROUP# STATUS o --- ---

o 1 CURRENT o 2 INACTIVE o 3 INACTIVE o

o SQL> alter system switch logfile;

o

o SQL> alter system switch logfile;

o

o SQL> select group#, status from v$log;

o

o GROUP# STATUS o --- --- o 1 INACTIVE o 2 INACTIVE o 3 CURRENT

o After making the last online redo log file the CURRENT one, drop the first online redo log:

o SQL> alter database drop logfile group 1;

o Database altered.

As a DBA, you should already be aware that if you are going to drop a logfile group, it cannot be the current logfile group. I have run into instances; however, where attempting to drop the logfile group resulted in the following error as a result of the logfile group having

an activestatus:

SQL> ALTER DATABASE DROP LOGFILE GROUP 1;

ALTER DATABASE DROP LOGFILE GROUP 1

*

ERROR at line 1:

ORA-01624: log 1 needed for crash recovery of instance TESTDB (thread 1)

ORA-00312: online log 1 thread 1: '<file_name>'

Easy problem to resolve. Simply perform a checkpoint on the database:

SQL> ALTER SYSTEM CHECKPOINT GLOBAL;

System altered.

SQL> ALTER DATABASE DROP LOGFILE GROUP 1;

Database altered.

o Re-create the dropped redo log group in ASM (and a different size if desired):

o SQL> alter database add logfile group 1 size 250m;

o

Database altered.

o After re-creating the online redo log group, force a log switch. The online redo log group just created should become the CURRENT one:

o SQL> select group#, status from v$log;

o

o GROUP# STATUS o --- --- o 1 UNUSED

o 2 INACTIVE o 3 CURRENT o

o SQL> alter system switch logfile;

o

o SQL> select group#, status from v$log;

o

o GROUP# STATUS o --- --- o 1 CURRENT o 2 INACTIVE

3 ACTIVE

o After creating the first online redo log group, loop back to drop / re-create the next online redo logfile until all logs are rebuilt in ASM.

o Verify all online redo logfiles have been created in ASM:

o SQL> select a.group#, a.member, b.bytes

o 2 from v$logfile a, v$log b where a.group# = b.group#;

o

o GROUP# MEMBER BYTES o --- --- --- o 1 +TESTDB_DATA1/testdb/onlinelog/group_1.259.598486831

262144000

o 2 +TESTDB_DATA1/testdb/onlinelog/group_2.260.598487179 262144000

o 3 +TESTDB_DATA1/testdb/onlinelog/group_3.258.598487365 262144000

o 1

+FLASH_RECOVERY_AREA/testdb/onlinelog/group_1.259.598486879 262144000

o 2

+FLASH_RECOVERY_AREA/testdb/onlinelog/group_2.257.598487225 262144000

o 3

+FLASH_RECOVERY_AREA/testdb/onlinelog/group_3.260.598487411 262144000

o

6 rows selected.

129. Perform the following steps to relocate the SPFILE from the local file system to an ASM disk group.

o Create a text-based initialization parameter file from the current binary SPFILE located on the local file system:

o SQL> CREATE PFILE='$ORACLE_HOME/dbs/initTESTDB.ora' o 2 FROM SPFILE='$ORACLE_HOME/dbs/spfileTESTDB.ora';

o

File created.

o Create new SPFILE in an ASM disk group:

o SQL> CREATE

SPFILE='+TESTDB_DATA1/TESTDB/spfileTESTDB.ora' o 2 FROM PFILE='$ORACLE_HOME/dbs/initTESTDB.ora';

o

File created.

o Shutdown the Oracle database:

o SQL> SHUTDOWN IMMEDIATE o Database closed.

o Database dismounted.

ORACLE instance shut down.

o Update the text-based init<SID>.ora file with the new location of the SPFILE in ASM:

$ echo "SPFILE='+TESTDB_DATA1/TESTDB/spfileTESTDB.ora'"

> $ORACLE_HOME/dbs/initTESTDB.ora

o Remove (actually rename) the old SPFILE on the local file system so that the new text-based init<SID>.ora will be used:

$ mv $ORACLE_HOME/dbs/spfileTESTDB.ora

$ORACLE_HOME/dbs/BACKUP_ASM.spfileTESTDB.ora o Open the Oracle database using the new SPFILE:

SQL> STARTUP

130. Verify that all database files have been created in ASM:

131. $ sqlplus "/ as sysdba"

132.

133. SQL> @dba_files_all 134.

135. Tablespace Name /

136. File Class Filename File Size Auto Next Max

137. --- ---- --- ---

138. APEX22

+TESTDB_DATA1/testdb/datafile/apex22.263.598482381 104,857,600 NO 0 0

139. EXAMPLE

+TESTDB_DATA1/testdb/datafile/example.264.598482345 157,286,400 YES 655,360 34,359,721,984

140. FLOW_1

+TESTDB_DATA1/testdb/datafile/flow_1.262.598482405 52,494,336 NO 0 0

141. SYSAUX

+TESTDB_DATA1/testdb/datafile/sysaux.267.598482213 419,430,400 YES 10,485,760 34,359,721,984

142. SYSTEM

+TESTDB_DATA1/testdb/datafile/system.269.598482099 608,174,080 YES 10,485,760 34,359,721,984

143. TEMP +TESTDB_DATA1/testdb/tempfile/temp.261.598485663 536,870,912 YES 262,144,000 34,359,721,984

144. UNDOTBS1

+TESTDB_DATA1/testdb/datafile/undotbs1.256.598482299 209,715,200 YES 5,242,880 34,359,721,984

145. USERS +TESTDB_DATA1/testdb/datafile/users.270.598481673 2,382,888,960 YES 1,310,720 34,359,721,984

146. [ CONTROL FILE ]

+TESTDB_DATA1/testdb/controlfile/backup.268.598481391 147. [ ONLINE REDO LOG ]

+FLASH_RECOVERY_AREA/testdb/onlinelog/group_1.259.598486879 262,144,000

148. [ ONLINE REDO LOG ]

+FLASH_RECOVERY_AREA/testdb/onlinelog/group_2.257.598487225 262,144,000

149. [ ONLINE REDO LOG ]

+FLASH_RECOVERY_AREA/testdb/onlinelog/group_3.260.598487411 262,144,000

150. [ ONLINE REDO LOG ]

+TESTDB_DATA1/testdb/onlinelog/group_1.259.598486831 262,144,000

151. [ ONLINE REDO LOG ]

+TESTDB_DATA1/testdb/onlinelog/group_2.260.598487179 262,144,000

152. [ ONLINE REDO LOG ]

+TESTDB_DATA1/testdb/onlinelog/group_3.258.598487365 262,144,000

153. --- 154. sum

6,044,581,888 155.

15 rows selected.

156. At this point, the target database is open with all of its datafiles, controlfiles, online redo logfiles, tempfiles, and SPFILE stored in ASM. If we wanted to remove the database files that were stored on the local file system (which are actually now RMAN copies), this could be done from an RMAN session. You could also then remove the old version of the controfile(s) that were stored on the local file system:

If this is a production database, it would be best practice to first backup the database files on the local disk before removing them!

157. RMAN> DELETE NOPROMPT FORCE COPY;

158.

159. allocated channel: ORA_DISK_1

160. channel ORA_DISK_1: sid=139 devtype=DISK 161.

162. List of Datafile Copies

163. Key File S Completion Time Ckp SCN Ckp Time Name

173. List of Control File Copies

174. Key S Completion Time Ckp SCN Ckp Time Name 175. --- - --- --- --- ----

176. 43 A 14-AUG-06 7937583 14-AUG-06 +TESTDB_DATA1/testdb/controlfile/backup.261.598482421 177.

178. List of Archived Log Copies

179. Key Thrd Seq S Low Time Name

193. datafile copy

filename=/u02/oradata/TESTDB/datafile/o1_mf_users_2fb4cqf4_.dbf recid=48 stamp=598482496

194. deleted datafile copy 195. datafile copy

filename=/u02/oradata/TESTDB/datafile/o1_mf_apex22_2ft4eswu_.dbf recid=49 stamp=598482496

196. deleted datafile copy 197. datafile copy

filename=/u02/oradata/TESTDB/datafile/o1_mf_flow_1_2fb4cegw_.dbf recid=50 stamp=598482496

198. deleted control file copy 199. control file copy

filename=+TESTDB_DATA1/testdb/controlfile/backup.261.598482421 recid=43 stamp=598482423

200. deleted archive log 201. archive log

filename=+FLASH_RECOVERY_AREA/testdb/archivelog/2006_08_14/thread_

1_seq_34.259.598482825 recid=48 stamp=598482824 202. deleted archive log

203. archive log

filename=+FLASH_RECOVERY_AREA/testdb/archivelog/2006_08_14/thread_

1_seq_35.258.598482825 recid=49 stamp=598482824 204. deleted archive log

205. archive log

filename=+FLASH_RECOVERY_AREA/testdb/archivelog/2006_08_14/thread_

1_seq_36.260.598482821 recid=47 stamp=598482824 206. Deleted 11 objects

207.

208. RMAN> exit 209.

210. $ rm /u02/oradata/TESTDB/controlfile/o1_mf_8du3s3er_.ctl 211. $ rm /u02/oradata/TESTDB/controlfile/o1_mf_y2is93je_.ctl

Migrating Oracle Database from ASM to Local File System

The following query lists database files as they exist in ASM for the TESTDB database. All of the files listed in this query will be relocated from ASM to the local file system:

$ ORACLE_SID=TESTDB; export ORACLE_SID

$ sqlplus "/ as sysdba"

SQL> @dba_files_all Tablespace Name /

File Class Filename File Size Auto Next Max

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

APEX22 +TESTDB_DATA1/testdb/datafile/apex22.263.598482381 419,430,400 YES 10,485,760 34,359,721,984

SYSTEM +TESTDB_DATA1/testdb/datafile/system.269.598482099 608,174,080 YES 10,485,760 34,359,721,984

TEMP +TESTDB_DATA1/testdb/tempfile/temp.261.598485663 536,870,912 YES 262,144,000 34,359,721,984

UNDOTBS1 +TESTDB_DATA1/testdb/datafile/undotbs1.256.598482299 209,715,200 YES 5,242,880 34,359,721,984

USERS +TESTDB_DATA1/testdb/datafile/users.270.598481673 2,382,888,960 YES 1,310,720 34,359,721,984

[ CONTROL FILE ] +TESTDB_DATA1/testdb/controlfile/backup.268.598481391 [ ONLINE REDO LOG ]

+FLASH_RECOVERY_AREA/testdb/onlinelog/group_1.259.598486879 262,144,000 [ ONLINE REDO LOG ]

+FLASH_RECOVERY_AREA/testdb/onlinelog/group_2.257.598487225 262,144,000 [ ONLINE REDO LOG ]

+FLASH_RECOVERY_AREA/testdb/onlinelog/group_3.260.598487411 262,144,000 [ ONLINE REDO LOG ] +TESTDB_DATA1/testdb/onlinelog/group_1.259.598486831

Also note that the target database uses an SPFILE which is stored in ASM. The target database starts using a text-based init<SID>.ora ($ORACLE_HOME/dbs/initTESTDB.ora) which defines the location of the SPFILE in ASM:

SPFILE='+TESTDB_DATA1/TESTDB/spfileTESTDB.ora'

Use the following steps to fully migrate an existing Oracle database from ASM to a local file system:

1. With the target database open, edit the initialization

parameter control_files and db_create_file_dest to point to locations on the local file system (/u02/oradata). Also configuredb_recovery_file_dest to point to the Flash Recovery Area on the local file system

(/u02/flash_recovery_area):

6. SQL> ALTER SYSTEM SET db_create_file_dest='/u02/oradata' SCOPE=spfile;

7.

8. System altered.

9.

10. SQL> ALTER SYSTEM SET

db_recovery_file_dest='/u02/flash_recovery_area' SCOPE=spfile;

11.

System altered.

12. Backup the current SPFILE (in ASM) to a text-based init<SID>.ora file on the local file system. Then convert the text-based init<SID>.ora file to a binary SPFILE on the local file system:

13. SQL> HOST mv $ORACLE_HOME/dbs/initTESTDB.ora

$ORACLE_HOME/dbs/BACKUP_ASM.initTESTDB.ora 14.

15. SQL> CREATE PFILE='$ORACLE_HOME/dbs/initTESTDB.ora' FROM SPFILE='+TESTDB_DATA1/TESTDB/spfileTESTDB.ora';

16.

17. File created.

18.

19. SQL> CREATE SPFILE='$ORACLE_HOME/dbs/spfileTESTDB.ora' FROM PFILE='$ORACLE_HOME/dbs/initTESTDB.ora';

20.

File created.

21. Startup the target database in NOMOUNT mode:

22. SQL> shutdown immediate 23. Database closed.

24. Database dismounted.

25. ORACLE instance shut down.

26.

27. SQL> startup nomount 28. ORACLE instance started.

29.

30. Total System Global Area 285212672 bytes 31. Fixed Size 1260420 bytes 32. Variable Size 150996092 bytes 33. Database Buffers 130023424 bytes

Redo Buffers 2932736 bytes

34. From an RMAN session, copy one of your controlfiles from ASM to its new location on the local file system. The new controlfile will be copied to the value specified in the initialization parametercontrol_files:

35. RMAN> RESTORE CONTROLFILE FROM

'+TESTDB_DATA1/TESTDB/CONTROLFILE/backup.268.598481391';

36.

37. Starting restore at 15-AUG-06

38. using target database control file instead of recovery catalog 39. allocated channel: ORA_DISK_1

40. channel ORA_DISK_1: sid=156 devtype=DISK 41.

42. channel ORA_DISK_1: copied control file copy 43. output filename=/u02/oradata/TESTDB/control01.ctl

Finished restore at 15-AUG-06

44. From an RMAN or SQL*Plus session, mount the database. This will mount the database using the controlfile stored on the local file system:

45. RMAN> ALTER DATABASE MOUNT;

46.

47. using target database control file instead of recovery catalog database mounted

48. From an RMAN session, copy the database files from ASM to the local file system:

49. RMAN> BACKUP AS COPY DATABASE FORMAT '/u02/oradata/TESTDB/%U';

50.

51. Starting backup at 15-AUG-06 52. allocated channel: ORA_DISK_1

53. channel ORA_DISK_1: sid=156 devtype=DISK 54. channel ORA_DISK_1: starting datafile copy 55. input datafile fno=00005

name=+TESTDB_DATA1/testdb/datafile/users.270.598481673

56. output filename=/u02/oradata/TESTDB/data_D-TESTDB_I-2370649665_TS-USERS_FNO-5_0vhqpltd tag=TAG20060815T101925 recid=51

stamp=598530181

57. channel ORA_DISK_1: datafile copy complete, elapsed time: 00:03:36 58. channel ORA_DISK_1: starting datafile copy

59. input datafile fno=00001

name=+TESTDB_DATA1/testdb/datafile/system.269.598482099

60. output filename=/u02/oradata/TESTDB/data_D-TESTDB_I-2370649665_TS-SYSTEM_FNO-1_10hqpm45 tag=TAG20060815T101925 recid=52

stamp=598530235

61. channel ORA_DISK_1: datafile copy complete, elapsed time: 00:00:55 62. channel ORA_DISK_1: starting datafile copy

63. input datafile fno=00003

name=+TESTDB_DATA1/testdb/datafile/sysaux.267.598482213

64. output filename=/u02/oradata/TESTDB/data_D-TESTDB_I-2370649665_TS-SYSAUX_FNO-3_11hqpm5s tag=TAG20060815T101925 recid=53

stamp=598530274

65. channel ORA_DISK_1: datafile copy complete, elapsed time: 00:00:46 66. channel ORA_DISK_1: starting datafile copy

67. input datafile fno=00002

name=+TESTDB_DATA1/testdb/datafile/undotbs1.256.598482299

68. output filename=/u02/oradata/TESTDB/data_D-TESTDB_I-2370649665_TS-UNDOTBS1_FNO-2_12hqpm7a tag=TAG20060815T101925 recid=54

stamp=598530304

69. channel ORA_DISK_1: datafile copy complete, elapsed time: 00:00:25 70. channel ORA_DISK_1: starting datafile copy

71. input datafile fno=00004

name=+TESTDB_DATA1/testdb/datafile/example.264.598482345

72. output filename=/u02/oradata/TESTDB/data_D-TESTDB_I-2370649665_TS-EXAMPLE_FNO-4_13hqpm83 tag=TAG20060815T101925 recid=55

stamp=598530323

73. channel ORA_DISK_1: datafile copy complete, elapsed time: 00:00:25 74. channel ORA_DISK_1: starting datafile copy

75. input datafile fno=00006

name=+TESTDB_DATA1/testdb/datafile/apex22.263.598482381

76. output filename=/u02/oradata/TESTDB/data_D-TESTDB_I-2370649665_TS-APEX22_FNO-6_14hqpm8s tag=TAG20060815T101925 recid=56

stamp=598530343

77. channel ORA_DISK_1: datafile copy complete, elapsed time: 00:00:16 78. channel ORA_DISK_1: starting datafile copy

79. input datafile fno=00007

name=+TESTDB_DATA1/testdb/datafile/flow_1.262.598482405

80. output filename=/u02/oradata/TESTDB/data_D-TESTDB_I-2370649665_TS-FLOW_1_FNO-7_15hqpm9c tag=TAG20060815T101925 recid=57

stamp=598530353

81. channel ORA_DISK_1: datafile copy complete, elapsed time: 00:00:07 82. channel ORA_DISK_1: starting datafile copy

83. copying current control file

84. output filename=/u02/oradata/TESTDB/cf_D-TESTDB_id-2370649665_16hqpm9j tag=TAG20060815T101925 recid=58 stamp=598530356

85. channel ORA_DISK_1: datafile copy complete, elapsed time: 00:00:01 86. channel ORA_DISK_1: starting full datafile backupset

87. channel ORA_DISK_1: specifying datafile(s) in backupset 88. including current SPFILE in backupset

89. channel ORA_DISK_1: starting piece 1 at 15-AUG-06 90. channel ORA_DISK_1: finished piece 1 at 15-AUG-06 91. piece handle=/u02/oradata/TESTDB/17hqpm9k_1_1

tag=TAG20060815T101925 comment=NONE

92. channel ORA_DISK_1: backup set complete, elapsed time: 00:00:02 Finished backup at 15-AUG-06

93. From an RMAN session, update the control file / data dictionary so that all database files point to the RMAN copy made on the local file system:

94. RMAN> SWITCH DATABASE TO COPY;

95.

96. datafile 1 switched to datafile copy "/u02/oradata/TESTDB/data_D-TESTDB_I-2370649665_TS-SYSTEM_FNO-1_10hqpm45"

97. datafile 2 switched to datafile copy "/u02/oradata/TESTDB/data_D-TESTDB_I-2370649665_TS-UNDOTBS1_FNO-2_12hqpm7a"

98. datafile 3 switched to datafile copy "/u02/oradata/TESTDB/data_D-TESTDB_I-2370649665_TS-SYSAUX_FNO-3_11hqpm5s"

99. datafile 4 switched to datafile copy "/u02/oradata/TESTDB/data_D-TESTDB_I-2370649665_TS-EXAMPLE_FNO-4_13hqpm83"

100. datafile 5 switched to datafile copy "/u02/oradata/TESTDB/data_D-TESTDB_I-2370649665_TS-USERS_FNO-5_0vhqpltd"

101. datafile 6 switched to datafile copy "/u02/oradata/TESTDB/data_D-TESTDB_I-2370649665_TS-APEX22_FNO-6_14hqpm8s"

datafile 7 switched to datafile copy "/u02/oradata/TESTDB/data_D-TESTDB_I-2370649665_TS-FLOW_1_FNO-7_15hqpm9c"

102. From a SQL*Plus session, perform incomplete recovery and open the database using the RESETLOGS option:

103. SQL> RECOVER DATABASE USING BACKUP CONTROLFILE UNTIL CANCEL;

104.

105. ORA-00279: change 7970332 generated at 08/15/2006 10:12:34 needed for thread 1

106. ORA-00289: suggestion :

107. /u02/flash_recovery_area/TESTDB/archivelog/2006_08_15/o1_mf_1_5_

%u_.arc

108. ORA-00280: change 7970332 for thread 1 is in sequence #5 109.

110.

111. Specify log: {=suggested | filename | AUTO | CANCEL}

112. CANCEL

113. Media recovery cancelled.

114.

115. SQL> ALTER DATABASE OPEN RESETLOGS;

116.

Database altered.

117. From a SQL*Plus session, re-create any tempfiles that are still

currently using ASM to the local file system. This is done by simply dropping the tempfiles from ASM and re-creating them in the local file system. This example relies on the initialization

parameter db_create_file_dest=/u02/oradata:

118. SQL> select tablespace_name, file_name, bytes from dba_temp_files;

119.

120. TABLESPACE_NAME FILE_NAME BYTES 121. --- --- --- 122. TEMP +TESTDB_DATA1/testdb/tempfile/temp.261.598485663

536870912 123.

124. SQL> alter database tempfile

125. 2 '+TESTDB_DATA1/testdb/tempfile/temp.261.598485663' 126. 3 drop including datafiles;

127.

128. Database altered.

129.

130. SQL> alter tablespace temp add tempfile size 512m 131. 2 autoextend on next 250m maxsize unlimited;

132.

133. Tablespace altered.

134.

135. SQL> select tablespace_name, file_name, bytes from dba_temp_files;

136.

137. TABLESPACE_NAME FILE_NAME BYTES 138. --- --- ---

TEMP /u02/oradata/TESTDB/datafile/o1_mf_temp_2g3spvq5_.tmp 536870912

If users are currently accessing the tempfile(s) you are attempting to drop, you may receive the following error:

SQL> alter database tempfile

2 '+TESTDB_DATA1/testdb/tempfile/temp.261.598485663' 3 drop including datafiles;

ERROR at line 1:

ORA-25152: TEMPFILE cannot be dropped at this time

As for the poor users who were using the tempfile, their transaction will end and will be greeted with the following error message:

SQL> @testTemp.sql

join dba_extents c on (b.segment_name = c.segment_name) *

ERROR at line 4:

ORA-00372: file 601 cannot be modified at this time ORA-01110: data file 601:

'+TESTDB_DATA1/testdb/tempfile/temp.261.598485663' ORA-00372: file 601 cannot be modified at this time ORA-01110: data file 601:

'+TESTDB_DATA1/testdb/tempfile/temp.261.598485663'

If this happens, you should attempt to drop the tempfile again so the operation is successful:

SQL> alter database tempfile

2 '+TESTDB_DATA1/testdb/tempfile/temp.261.598485663' 3 drop including datafiles;

Database altered.

139. From a SQL*Plus session, re-create any online redo logfiles that are still currently using ASM to the local file system. This is done by simply dropping the logfiles from ASM and re-creating them on the local file system.

This example relies on the initialization

parameters db_create_file_dest=/u02/oradata and db_recovery_file_dest=/u 02/flash_recovery_area:

o Determine the current online redo logfiles to move to the local file system by examining the file names (and sizes) from V$LOGFILE:

o SQL> select a.group#, a.member, b.bytes

o 2 from v$logfile a, v$log b where a.group# = b.group#;

o

o GROUP# MEMBER BYTES o --- --- --- o 1 +TESTDB_DATA1/testdb/onlinelog/group_1.259.598486831

262144000

o 2 +TESTDB_DATA1/testdb/onlinelog/group_2.260.598487179 262144000

o 3 +TESTDB_DATA1/testdb/onlinelog/group_3.258.598487365 262144000

o 1

+FLASH_RECOVERY_AREA/testdb/onlinelog/group_1.259.598486879 262144000

o 2

+FLASH_RECOVERY_AREA/testdb/onlinelog/group_2.257.598487225 262144000

o 3

+FLASH_RECOVERY_AREA/testdb/onlinelog/group_3.260.598487411 262144000

o

6 rows selected.

o Force a log switch until the last redo log is marked "CURRENT" by issuing the following command:

o SQL> select group#, status from v$log;

o

o GROUP# STATUS o --- --- o 1 CURRENT o 2 INACTIVE o 3 INACTIVE o

o SQL> alter system switch logfile;

o

o SQL> alter system switch logfile;

o

o SQL> select group#, status from v$log;

o

o GROUP# STATUS o --- --- o 1 INACTIVE o 2 INACTIVE o 3 CURRENT

o After making the last online redo log file the CURRENT one, drop the first online redo log:

o SQL> alter database drop logfile group 1;

o

Database altered.

As a DBA, you should already be aware that if you are going to drop a logfile group, it cannot be the current logfile group. I have run into instances; however, where attempting to drop the logfile group resulted in the following error as a result of the logfile group having

an activestatus:

SQL> ALTER DATABASE DROP LOGFILE GROUP 1;

ALTER DATABASE DROP LOGFILE GROUP 1

ALTER DATABASE DROP LOGFILE GROUP 1

In document Oracle DBA (Page 117-163)