BLOCK CORRUPTION (ORACLE8 에서 ORA-1578 조치방법)

제품 : ORACLE SERVER
작성날짜 : 2002-04-26
BLOCK CORRUPTION (ORACLE8 에서 ORA-1578 조치방법)
================================================
PURPOSE
본 bulletin 에서는 Oracle8 에서의 data block corruption error
(ORA-1578) 에 대해서 조치방법을 알아본다.
BACKGROUND
ORA-1578 은 data block 에 corruption 이 생긴 경우에 발생한다.
이 error 는 corruption 이 발생한 곳의 file number 와 block
number 를 알려준다.
본 bulletin 에서는 file number 를 'f', block number 를 'b' 로
지칭하기로 한다. ORA-1578 과 함께 return 되는 file number는
relative file number 가 아닌 absolute file number 를 의미한다.
(Oracle8 에서 새롭게 소개된 relative file number 에 대해서는 본
bulletin 의 6번 항목에서 다루도록 한다.)
RESOLUTION
1. 최선의 해결책은 backup 받아둔 file 을 restore 한 후
recover 작업을 하는 것이다. 이 작업을 위해서는 archive
log mode 로 운영 중이어야 한다.
dba_data_files 또는 v$datafile 과 ORA-1578 error 에서
return 된 absolute file number 를 이용하여 corruption이
발생한 datafile 의 이름을 알아낼 수 있다.
Oracle8 에서는, dba_data_files 에 absolute file number
(file_id) 와 함께 relative file number (relative_fno) 를
가지고 있다.
v$datafile 에는 아직 absolute file number (file#) 만을 가진다.
backup 을 restore 하기 전에 hardware 의 문제를 fix 해야 할
필요가 있을 수 있다. 만약 corruption 이 disk 불량에 의해
발생하였다면, backup 받아둔 datafile 을 문제가 없는 disk 에
restore 하고, startup mount 한 후, 새로운 위치로 datafile
rename 을 하고 recover 한다.
만약 해당 datafile 이 system tablespace 에 속하지 않는다면
offline tablespace recovery 도 가능하다.
2. backup datafile 을 restore 하고 recover 하지 않을 것이라면
우선, 어떤 object 에서 corruption 이 발생하였는지 확인해야
한다.
다음의 SQL 문을 이용한다.
SELECT owner, segment_name, segment_type, relative_fno
FROM dba_extents
WHERE file_id = f
AND b BETWEEN block_id AND block_id + blocks - 1 ;
3. 해당 segment 가 non-data dictionary index라면, 해당 index를
drop 한 후 재생성한다.
4. 해당 segment 가 table 이라면, corruption 이 발생한 block 의
data 는 소실된 것이다.
5. 만약 해당 table 에 대한 최근의 export dump file이 존재한다면,
해당 table 을 drop 한 후 import 함으로써 복구할 수 있다.
최근의 export dump file 이 없거나 이를 export 받을 수 있는
최근의 backup 이 없다면 다음과 같은 방법을 이용할 수 있다.
6. corruption 이 발생한 non-clustered table 에서 corrupted
block 을 access 하지 않고 나머지 data 들을 select 할 수
있도록 ROWID 를 이용할 수 있다. non-clustered table의 모든
row 들은 해당 row 의 물리적인 주소를 가리키는 고유한 ROWID를
갖는다(해당 row가 여러 block에 조각이 나 있다면 첫번째 조각에
대한 주소). clustered table 인 경우에는 서로 다른 table 의
data 들이 같은 block 에 존재할 수 있으며, 같은 ROWID 를 가질
수 있다.
Oracle 은 index 를 구성하기 위하여 내부적으로 ROWID 를 사용한다.
따라서 where 절에 ROWID 를 사용하여 select 하면 강제로 range
scan 를 할 수 있다.
Oracle8 에서는 이를 위하여 ROWID hint 를 사용할 필요가
없어졌다.
ROWID 를 이용하여 table 로부터 data 를 추출하기
예제로써 ACCT_NO, PERSON, WEEKNO 등의 column 으로 구성된
SCOTT user의 EXAMPLE 이라는 table 이 있다고 가정한다.
이때, ORA-1578 "ORACLE data block corrupted (file # 5,
block # 2) 가
발생하였다고 가정하자.
ROWID in Oracle8 :
Oracle8 에서의 ROWID 는 'OOOOOOFFFBBBBBBSSS' format 을 가진다.
OOOOOO = data object number
Oracle7 의 object_id 와는 별도로 segment 의 id 를 갖는다.
dba_objects 의 data_object_id 에서 확인가능
FFF = relative file number
BBBBBB = block number
SSS = row number
Oracle8에서의 ROWID 는 absolute file number 가 아닌 relative
file number 를 갖는다는 점에 주목해야 한다.
relative file number 는 tablespace 에 대해 상대적이며
(tablespace마다 첫번째, 두번째, 세번째 datafile 을 가질 수
있음을 의미) absolute file number 는 전체 database 내에서
고유하다. 두개의 서로다른 file 들이 동일한 relative file
number를 가질 수 있다.
만약 EXAMPLE table 에서 ACCT_NO, ROWID 를 select하면 다음과
같은 결과가 나올 수 있다.
ACCT_NO ROWID
12345 AAAAh3AAGAAACJAAAA
19283 AAAAh3AAGAAACJAAAB
22345 AAAAh4AAFAAAAADAAA
60372 AAAAh4AAFAAAAADAAB
33456 AAAAh5AAEAAAAIuAAA
29473 AAAAh5AAEAAAAIuAAB
이러한 format 을 extended ROWID character format 이라고
지칭한다.
extended ROWID 는 number 이므로 substr 함수를 이용하여
extended ROWID 로부터 일부를 떼어낼 수 없다.
ROWID 를 생성하기 위해서는 모든 component 를 알아야 한다.
그런다음, DBMS_ROWID package 의 function 을 이용하여 ROWID 를
생성할 수 있다.
function rowid_create(rowid_type IN number,
object_number IN number,
relative_fno IN number,
block_number IN number,
row_number IN number)
return rowid;
pragma RESTRICT_REFERENCES(rowid_create,WNDS,RNDS,WNPS,RNPS);
-- rowid_type - type (restricted=0/extended=1)
-- object_number - data object number
(rowid_object_undefined for )
-- restricted
-- relative_fno - relative file number
-- block_number - block number in this file
-- file_number - file number in this block
corruption 이 발생한 block 의 data object number 를 알기
위해서는 dba_objects 를 조회한다.
SELECT data_object_id FROM dba_objects
WHERE owner = 'SCOTT'
AND object_name = 'EXAMPLE' ;
DATA_OBJECT_ID
2168
우리는 이미, 위에서 사용한 다음의 SQL 에 의해서 relative file
number 를 알고 있다.
SELECT owner, segment_name, segment_type, relative_fno
FROM dba_extents
WHERE file_id = 5
AND 2 BETWEEN block_id AND block_id + blocks - 1 ;
OWNER SEGMENT_NAME SEGMENT_TYPE RELATIVE_F
SCOTT EXAMPLE TABLE 5
corruption 이 발생한 block 이 2이므로, block# 2 이전의 access
가능한 마지막 ROWID 는 block# 1 에 존재한다. 그리고 block# 2
이후의 access 가능한 첫 ROWID 는 block# 3에 존재한다. block
안에 몇개의 row number 가 존재할 지 모르므로 row number 0 을
이용한다.
이제 corrupted block 이전의 ROWID 와 이후의 ROWID 를 생성할
준비가 끝났다.
corrupted block 이전의 ROWID :
SELECT DBMS_ROWID.ROWID_CREATE(1,2168,5,1,0) FROM example ;
DBMS_ROWID.ROWID_C
AAAAh4AAFAAAAABAAA
corrupted block 이후의 ROWID :
SELECT DBMS_ROWID.ROWID_CREATE(1,2168,5,3,0) FROM example ;
DBMS_ROWID.ROWID_C
AAAAh4AAFAAAAADAAA
다음으로, EXAMPLE table 과 같은 spec 으로 table 을 하나 만든다.
CREATE TABLE temp AS SELECT * FROM example WHERE 1=2;
그리고 corrupted block 이외의 block 에서 data 를 추출하여
insert 한다.
INSERT INTO temp SELECT * FROM example
WHERE ROWID <= 'AAAAh4AAFAAAAABAAA';
INSERT INTO temp SELECT * FROM example
WHERE ROWID >= 'AAAAh4AAFAAAAADAAA';
이후 원본 table 을 drop 하고, TEMP 를 rename 한다.
7. 만약 data dictionary 에 속하는 table, index 또는 rollback
segment에 corrupted block 이 발생하였다면 Oracle Support 의
지원을 받는다.
8. 일반적으로, ORA-1578 은 hardware 의 문제때문에 유발된다.
하지만 만약에 ORA-600[3374] 가 발생한다면 memory 상에서
corruption 이 발생한 경우이다.
이 경우 database 를 restartup 하면 문제가 해결될 수 있다.
Reference Documents
--------------------

Hello,
Please check the link i posted.
Example: Detecting Corruption
The CHECK_OBJECT procedure checks the specified object, and populates the repair table with information about corruptions and repair directives. You can optionally specify a range, partition name, or subpartition name when you want to check a portion of an object.
Validation consists of checking all blocks in the object that have not previously been marked corrupt. For each block, the transaction and data layer portions are checked for self consistency. During CHECK_OBJECT, if a block is encountered that has a corrupt buffer cache header, then that block is skipped.
The following is an example of executing the CHECK_OBJECT procedure for the scott.dept table.
SET SERVEROUTPUT ON
DECLARE num_corrupt INT;
BEGIN
num_corrupt := 0;
DBMS_REPAIR.CHECK_OBJECT (
SCHEMA_NAME => 'SCOTT',
OBJECT_NAME => 'DEPT',
REPAIR_TABLE_NAME => 'REPAIR_TABLE',
CORRUPT_COUNT => num_corrupt);
DBMS_OUTPUT.PUT_LINE('number corrupt: ' || TO_CHAR (num_corrupt));
END;
SQL*Plus outputs the following line, indicating one corruption:
number corrupt: 1

Similar Messages

  • LAST ORA-1578 RECOVERY METHOD

    제품 : ORACLE SERVER
    작성날짜 : 2002-04-12
    LAST ORA-1578 RECOVERY METHOD
    =============================
    Purpose
    Ora-1578 error 에 대한 조치 방법에 대해 이해하도록 합니다.
    Problem Description
    일반적으로
    ORA-1578 : ORACLE data block corrupted (file # num,block # num)
    은 해당 objects를 drop하고 recreate하여 처리할 수 있읍니다.
    하지만, backup이 안되어 있거나 되어 있더라도 시점에 문제가 생기면,
    당황할 수 밖에 없습니다.
    Solution Description
    이 자료에서는 손상된 db block만을 제외한 다른 block들을 recover하는
    방법을 알려 드립니다.
    1. 손상된 table과 똑같은 spec의 new table생성
    2. 아래의 pl/sql 실행
    ** This program uses a cursor to select the ROWID from INDEX
    ** for the corrupted table. And insert the all of the normal rows
    ** into the new table.
    DECLARE
    CURSOR c1 is
    SELECT ROWID FROM corrupted_table
    WHERE indexed_column > smallest_value_of_the_indexes
    AND SUBSTR(ROWID,1,8) != '000005DA';
    /* Here,the value of '000005DA' is displayed in decimal
    when ORA-1578 error happens. */
    rowid_value CHAR(18);
    count_value NUMBER;
    BEGIN
    OPEN c1;
    FOR i IN 1..total_number_of_the_rows_from_corrupted_table LOOP
    FETCH c1 INTO rowid_value;
    EXIT WHEN c1%NOTFOUND;     
    INSERT INTO new_table
    SELECT * FROM corrupted_table WHERE ROWID = rowid_value;
    count_value := count_value + 1;
    IF count_value = 10000 THEN /* Let's think commit per 10000 rows */
    COMMIT;
    count_value := 0;
    END-IF;
    END LOOP;
    CLOSE c1;
    END;
    3. new table의 data건수등 확인후에 기존 table을 다른 이름으로
    rename한 후 new table을 original table name으로 rename
    SQL> rename corrupted_table to table_save;
    SQL> rename new_table to corrupted_table;
    4. 필요한 index생성이나 constraint enable
    5. 손상된 row에 대한 내용은,
    최소한 primary key나 indexed columns에 대한 내용을
    SQL> spool chk.log
    SQL> select rowid,indexed_column1,indexed_column2,indexed_column3...
    from corrupted_table
    where indexed_column1 > smallest_value_of_the_index;
    SQL> spool off
    를 통하여 chk.log에 저장후,
    SQL> select count(*) from corrupted_table;
    혹은 table export등을 통하여 확인된 rowid의 block number를 비교하여
    보면 됩니다.
    주의사항은 여기에서 ora-1578과 동반되는 block number는 decimal이니,
    hexa로 환산해야 합니다.

    you did not try to recover the datafile until get syncronyns with the controlfile ?
    RECOVER
    RECOVER {general | managed | END BACKUP}
    where the general clause has the following syntax:
    [AUTOMATIC] [FROM location]
    { {full_database_recovery | partial_database_recovery |LOGFILE filename}
    [ {TEST | ALLOW integer CORRUPTION } [TEST | ALLOW integer CORRUPTION ]...]
    |CONTINUE [DEFAULT]|CANCEL}
    where the full_database_recovery clause has the following syntax:
    [STANDBY] DATABASE
    [ {UNTIL {CANCEL | TIME date | CHANGE integer} | USING BACKUP CONTROLFILE}
    [UNTIL {CANCEL | TIME date | CHANGE integer} | USING BACKUP CONTROLFILE]...]
    where the partial_database_recovery clause has the following syntax:
    {TABLESPACE tablespace [, tablespace]... | DATAFILE datafilename [, datafilename]...
    | STANDBY
    {TABLESPACE tablespace [, tablespace]... | DATAFILE datafilename [, datafilename]...}
    UNTIL [CONSISTENT] [WITH] CONTROLFILE }
    where the managed clause has the following syntax:
    MANAGED STANDBY DATABASE
    [ {NODELAY | [TIMEOUT] integer | CANCEL [IMMEDIATE] [NOWAIT]}
    | [DISCONNECT [FROM SESSION] ] [FINISH [NOWAIT] ] ]
    http://download-west.oracle.com/docs/cd/B10501_01/server.920/a90842/ch13.htm#1010523
    Joel P�rez

  • ORA-1578 조치를 위한 EVENT 10231

    제품 : ORACLE SERVER
    작성날짜 : 2000-08-28
    ORA-1578 조치를 위한 event 10231
    ========================================================================
    table access 시 ora-1578이 발생하는 경우, 이미 다른 bulletin(10062, 11885)에
    정리된 것과 같이 corrupt된 block을 제외한 나머지 block의 data를 index
    search를 통해 조회하여 다른 table로 copy하는 방법이 사용된다.
    그런데, 만약 ora-1578을 만난 table에 index가 전혀 없다면 이 자료에서 설명할
    event 10231을 사용할 수 있다.
    이 event는 V7, V8 모두 적용 가능하다.
    1. event 10231의 효과
    ~~~~~~~~~~~~~~~~~~~~~
    event 10231은 full table scan 시에 corrupt된 block을 skip할 수 있도록
    해주는 event이다. block corruption을 나타내는 ora-1578의 조치 방법으로
    이용될 수 있으나, 이 event가 효과를 발휘하는 corruption의 형태가 모든 것을
    포함하는 것은 아니기 때문에 이 event를 사용하여도 여전히 ora-1578이 발생할 수
    있다.
    대부분의 ora-1578에 대해서 corrupt된 block을 skip하는 것이 확인되었으나,
    적용되지 않는 경우에 대해서 오라클이 보장하지는 않는다.
    앞에서 설명한 바와 같이 이 event는 full table scan 시에 효과가 있으므로,
    ora-1578이 발생하면 이것을 이용하여 corrupt된 block을 포함하는 table을
    export 받아내거나, "create table as select" 구문을 이용하여 다른 table로
    data를 copy할 수 있다.
    corrupted된 block 내의 data는 잃게 된다.
    Oracle 7.1.X version까지는 이 event가 물리적인 corruption을 제외한 'soft
    corrupt' block에만 효과가 있었으나, Oracle 7.2 이상부터는 (V8포함) media
    corruption을 포함한 다양한 형태의 corruption에 대해서 효과를 가지게 되어
    더욱 유용해졌다.
    그러나, 여전히 모든 case를 cover하는 보장은 되지 못한다.
    [주의] 이 event는 data dictionary table에 대해서는 사용할 수 없다.
    2. event 설정 및 사용 방법
    ~~~~~~~~~~~~~~~~~~~~~~~~~~
    이 event는 disk로부터 읽는 block에만 효과가 있다. corrupt된 block이 memory,
    즉, cache buffer 내에 이미 올라와 있다면 바라는 효과를 얻을 수 없기 때문에
    일단 이 parameter를 사용하기 전에는 db를 shutdown시키는 것이 권고된다.
    event를 사용하는 절차를 정리하면 다음과 같다.
    (1) database shutdown
    (2) cd $ORACLE_HOME/dbs
    initSID.ora (SID는 env | grep SID하여 확인) file 내에 다음과 같이
    event를 지정하는 parameter를 추가한다. 위치는 관계없다.
    event="10231 trace name context forever, level 10"
    (3) 다음과 같이 restrict mode로 startup하고, parameter가 설정되었는지
    확인한다. startup 시 parameter error가 발생하면 위의 지정에서
    콤마와 같은 부분이 문제가 있는지 다시 한번 확인한다.
    SVRMGR>startup restrict;
    SVRMGR>show parameter event
    NAME TYPE VALUE
    event string 10231 trace name context forev
    (4) full table scan operation을 이용하여 table로 부터 data를 꺼낸다.
    예를 들어 scott user의 dept에서 ora-1578이 발생한 것을 가정하고,
    다음과 같은 조치가 가능하다.
    <Example 1>
    os> exp scott/tiger file=test.dmp tables=dept
    <Example 2>
    SQL> create table test as select * from dept;
    3. data 발췌 후 작업
    ~~~~~~~~~~~~~~~~~~~~
    일단 export나 create table as select 문장의 수행으로 인해 data를 얻었으면,
    다음과 같이 작업한다.
    (1) initSID.ora 화일에서 event를 지정한 line을 제거한다.
    (2) db를 shutdown시키고 다시 startup시킨다.
    (3) show parameter event 를 수행하여 제거되었는지 확인한다.
    (4) corrupt된 table을 drop시키거나 rename시키고, table을 재생성한다.
    즉, 앞에서 예를 든 경우의 후속 작업은 각각 다음과 같다.
    <Example 1>
    SQL> drop table dept; (혹은 SQL>rename table dept to test;)
    os> imp scott/tiger file=test.dmp tables=dept
    이 때, space에 여유가 있다면 drop보다는 rename한 후 나중에
    import 결과를 확인하고 drop하는 것을 권고한다.
    단, rename을 하게 되면 index 등은 rename이 안 되므로 import 시
    index, constraint 등은 따로 생성하여야 한다.
    <Example 2>
    SQL> drop table dept;
    SQL> rename test to dept;
    (5) import 전에 rename한 경우나 create table as select를 이용하면,
    index, constraint, trigger 등은 재생성하여야 한다.
    또한, 필요한 권한을 grant하여 준다.

  • BLOCK CORRUPTION (ORA-1578) 처리 (ORACLE 8I NEW FEATURE)

    제품 : ORACLE SERVER
    작성날짜 : 2002-05-31
    BLOCK CORRUPTION (ORA-1578) 처리 (ORACLE 8I NEW FEATURE)
    ========================================================
    PURPOSE
    Block Corruption의 처리 방안에 대해 알아본다.
    Problem Description
    block corruption 시 10210, 10211,10231 의 event 를 사용해서
    해당 block 을 skip 할 수도 있지만 V8.1 이상에서는
    dbms_repair.fix_corrupt_blocks ,
    dbms_repair.skip_corrupt_block 를 이용하여
    corrupt가 발생한 block을 detect하고 skip, 또는 repair해주는 방안이
    제시되고 있다.
    Workaround
    Solution Description
    - 먼저 detecting 을 위해 db_block_checking =true 를 init.ora 에 set
    - dbms_repair 의 package 를 사용하는데 이 package 는 dbmsrpr.sql,
    prvtrpr.plb를 수행한다 .
    - sys 로 접속하여 package 를 실행한다.
    다음의 예제를 살펴보자
    T1 테이블에 corrupt 된 block 이 있다고 가정한다.
    SQL> desc t1
    Name Null? Type
    COL1 NOT NULL NUMBER(38)
    COL2 CHAR(512)
    SQL> analyze table t1 validate structure;
    analyze table t1 validate structure
    ERROR at line 1:
    ORA-01498: block check failure - see trace file
    이때 ANALYZE로 부터 발생된 trace file 에 corrupt 된 block 에 3 row 의
    (nrows = 3) data 가 있음을 알수 있다고 가정하자.
    DBMS_REPAIR.ADMIN_TABLES (repair and orphan key)
    ================================================
    ADMIN_TABLES 은 table 을 위한 repair table과,인덱스를 위한 orphan key
    tables을 제공한다.
    SQL> @adminCreate
    SQL> connect sys/change_on_install
    Connected.
    SQL>
    SQL> -- Repair Table
    SQL>
    SQL> declare
    2 begin
    3 -- Create repair table
    4 dbms_repair.admin_tables (
    5 -- table_name => 'REPAIR_TABLE',
    6 table_type => dbms_repair.repair_table,
    7 action => dbms_repair.create_action,
    8 tablespace => 'USERS'); -- default TS of SYS if not specified
    9 end;
    10 /
    PL/SQL procedure successfully completed.
    SQL> select owner, object_name, object_type
    2 from dba_objects
    3 where object_name like '%REPAIR_TABLE';
    OWNER OBJECT_NAME OBJECT_TYPE
    SYS DBA_REPAIR_TABLE VIEW
    SYS REPAIR_TABLE TABLE
    SQL>
    SQL> -- Orphan Key Table
    SQL>
    SQL> declare
    2 begin
    3 -- Create orphan key table
    4 dbms_repair.admin_tables (
    5 table_type => dbms_repair.orphan_table,
    6 action => dbms_repair.create_action,
    7 tablespace => 'USERS'); -- default TS of SYS if not specified
    8 end;
    9 /
    PL/SQL procedure successfully completed.
    SQL> select owner, object_name, object_type
    2 from dba_objects
    3 where object_name like '%ORPHAN_KEY_TABLE';
    OWNER OBJECT_NAME OBJECT_TYPE
    SYS DBA_ORPHAN_KEY_TABLE VIEW
    SYS ORPHAN_KEY_TABLE TABLE
    DBMS_REPAIR.CHECK_OBJECT
    =========================
    CHECK_OBJECT procedure 는 기술된 object를 check 하고, repair 를 위한 정보를 수집하기 위함이다.
    SQL> @checkObject
    SQL> set serveroutput on
    SQL>
    SQL> declare
    2 rpr_count int;
    3 begin
    4 rpr_count := 0;
    5 dbms_repair.check_object (
    6 schema_name => 'SYSTEM',
    7 object_name => 'T1',
    8 repair_table_name => 'REPAIR_TABLE',
    9 corrupt_count => rpr_count);
    10 dbms_output.put_line('repair count: ' || to_char(rpr_count));
    11 end;
    12 /
    repair count: 1
    PL/SQL procedure successfully completed.
    SQL> desc repair_table
    Name Null? Type
    OBJECT_ID NOT NULL NUMBER
    TABLESPACE_ID NOT NULL NUMBER
    RELATIVE_FILE_ID NOT NULL NUMBER
    BLOCK_ID NOT NULL NUMBER
    CORRUPT_TYPE NOT NULL NUMBER
    SCHEMA_NAME NOT NULL VARCHAR2(30)
    OBJECT_NAME NOT NULL VARCHAR2(30)
    BASEOBJECT_NAME VARCHAR2(30)
    PARTITION_NAME VARCHAR2(30)
    CORRUPT_DESCRIPTION VARCHAR2(2000)
    REPAIR_DESCRIPTION VARCHAR2(200)
    MARKED_CORRUPT NOT NULL VARCHAR2(10)
    CHECK_TIMESTAMP NOT NULL DATE
    FIX_TIMESTAMP DATE
    REFORMAT_TIMESTAMP DATE
    SQL> select object_name, block_id, corrupt_type, marked_corrupt,
    2 corrupt_description, repair_description
    3 from repair_table;
    OBJECT_NAME BLOCK_ID CORRUPT_TYPE MARKED_COR
    CORRUPT_DESCRIPTION
    REPAIR_DESCRIPTION
    T1 3 1 FALSE
    kdbchk: row locked by non-existent transaction
    table=0 slot=0
    lockid=32 ktbbhitc=1
    mark block software corrupt
    Data Extraction
    ===============
    repair table에 의하면 file 6 ,block 3 에 corrupt 이 났음을 알수 있다
    그러나 아직 이 block 은 corrupt 로 mark 되어 있지 않으므로 필요 data 를
    추출하여야 한다.
    1. ALTER SYSTEM DUMP (nrows = 3) 에 의해 block안에 있는 row수를 결정한다.
    2. corrupt object를 select 하여 가능한 정보를 추출한다.
    SQL> -- The following query can be used to salvage data from a corrupt block.
    SQL> -- Creating a temporary table facilitates data insertion.
    SQL> create table temp_t1 as
    2 select * from system.t1
    3 where dbms_rowid.rowid_block_number(rowid) = 3
    4 and dbms_rowid.rowid_to_absolute_fno (rowid, 'SYSTEM','T1') = 6;
    Table created.
    SQL> select col1 from temp_t1;
    COL1
    2
    3
    DBMS_REPAIR.FIX_CORRUPT_BLOCKS (ORA-1578)
    ============================================
    FIX_CORRUPT_BLOCKS procedure는 repair table 의 정보를 이용하여 corrupt
    blocks 을 fix 한다
    그러나 아직 full table scan 시 여전히 error 가 발생한다
    SQL> declare
    2 fix_count int;
    3 begin
    4 fix_count := 0;
    5 dbms_repair.fix_corrupt_blocks (
    6 schema_name => 'SYSTEM',
    7 object_name => 'T1',
    8 object_type => dbms_repair.table_object,
    9 repair_table_name => 'REPAIR_TABLE',
    10 fix_count => fix_count);
    11 dbms_output.put_line('fix count: ' || to_char(fix_count));
    12 end;
    13 /
    fix count: 1
    PL/SQL procedure successfully completed.
    SQL> select object_name, block_id, marked_corrupt
    2 from repair_table;
    OBJECT_NAME BLOCK_ID MARKED_COR
    T1 3 TRUE
    SQL> select * from system.t1;
    select * from system.t1
    ERROR at line 1:
    ORA-01578: ORACLE data block corrupted (file # 6, block # 3)
    ORA-01110: data file 6: '/tmp/ts_corrupt.dbf'
    DBMS_REPAIR.DUMP_ORPHAN_KEYS
    ==============================
    DUMP_ORPHAN_KEYS는 corrupt data 에 해당하는 index 를 나타내 준다
    SQL> select index_name from dba_indexes
    2 where table_name in (select distinct object_name from repair_table);
    INDEX_NAME
    T1_PK
    SQL> @dumpOrphanKeys
    SQL> set serveroutput on
    SQL>
    SQL> declare
    2 key_count int;
    3 begin
    4 key_count := 0;
    5 dbms_repair.dump_orphan_keys (
    6 schema_name => 'SYSTEM',
    7 object_name => 'T1_PK',
    8 object_type => dbms_repair.index_object,
    9 repair_table_name => 'REPAIR_TABLE',
    10 orphan_table_name => 'ORPHAN_KEY_TABLE',
    11 key_count => key_count);
    12 dbms_output.put_line('orphan key count: ' || to_char(key_count));
    13 end;
    14 /
    orphan key count: 3
    PL/SQL procedure successfully completed.
    SQL> desc orphan_key_table
    Name Null? Type
    SCHEMA_NAME NOT NULL VARCHAR2(30)
    INDEX_NAME NOT NULL VARCHAR2(30)
    IPART_NAME VARCHAR2(30)
    INDEX_ID NOT NULL NUMBER
    TABLE_NAME NOT NULL VARCHAR2(30)
    PART_NAME VARCHAR2(30)
    TABLE_ID NOT NULL NUMBER
    KEYROWID NOT NULL ROWID
    KEY NOT NULL ROWID
    DUMP_TIMESTAMP NOT NULL DATE
    SQL> select index_name, count(*) from orphan_key_table
    2 group by index_name;
    INDEX_NAME COUNT(*)
    T1_PK 3
    Note: orphan key table의 index 는 다시 rebuild 되어야 한다.
    DBMS_REPAIR.SKIP_CORRUPT_BLOCKS
    ===============================
    SKIP_CORRUPT_BLOCKS 은 table 과 index 의 corrupt block 을 skip 하는 것을 enable/disable 을 실시한다.
    Suggestion: SKIP_CORRUPT_BLOCKS 가 enabled되면 orphan key table의 모든
    index 는 모두 rebuild 되어야 한다. ( all index associated with object
    if DUMP_ORPHAN_KEYS was omitted).
    SQL> @skipCorruptBlocks
    SQL> declare
    2 begin
    3 dbms_repair.skip_corrupt_blocks (
    4 schema_name => 'SYSTEM',
    5 object_name => 'T1',
    6 object_type => dbms_repair.table_object,
    7 flags => dbms_repair.skip_flag);
    8 end;
    9 /
    PL/SQL procedure successfully completed.
    SQL> select table_name, skip_corrupt from dba_tables
    2 where table_name = 'T1';
    TABLE_NAME SKIP_COR
    T1 ENABLED
    SQL> -- rows in corrupt block skipped, no errors on full table scan
    SQL> select * from system.t1;
    COL1 COL2
    4 dddd
    5 eeee
    --> Notice the pk index has not yet been corrected.
    SQL> insert into system.t1 values (1,'aaaa');
    insert into system.t1 values (1,'aaaa')
    SQL> select * from system.t1 where col1 = 1;
    no rows selected
    DBMS_REPAIR.REBUILD_FREELISTS
    ===============================
    REBUILD_FREELISTS rebuilds freelists for the specified object.
    SQL> declare
    2 begin
    3 dbms_repair.rebuild_freelists (
    4 schema_name => 'SYSTEM',
    5 object_name => 'T1',
    6 object_type => dbms_repair.table_object);
    7 end;
    8 /
    PL/SQL procedure successfully completed.
    Rebuild Index
    =============
    Note: Every index identified in the orphan key table should be rebuilt to
    ensure consistent results.
    SQL> alter index system.t1_pk rebuild online;
    Index altered.
    SQL> insert into system.t1 values (1, 'aaaa');
    1 row created.
    SQL> select * from system.t1;
    COL1 COL2
    4 dddd
    5 eeee
    1 aaaa
    Reference Document
    ------------------

    Try look to alert<SID>.log file for full error report (you could paste it here).
    Also from alert log you could get real values for db_block_buffers and shared_pool_size parameters that used during instance startup.

  • DATA BLOCK CORRUPTION : ORA-1578 해결 방법

    제품 : ORACLE SERVER
    작성날짜 : 2005-08-18
    DATA BLOCK CORRUPTION : ORA-1578 해결 방법
    =========================================
    모든 오라클 데이타 블럭은 sequence 번호(seq)와 incarnation 번호(inc)를
    갖고 있다. ORA-1578 에러는 seq=0 이고 inc <> 0 (새로운 블럭이 아님)일 때
    발생한다. ORA-1578 에러는 ORA-600[3339] 에러와 함께 발생하곤 한다.
    ORA-1578 에러가 발생하면 corruption이 발생한 화일 번호와 블럭 번호를
    알려준다. 여기서는 이 때의 화일번호를 f, 블럭번호를 b 라고 부르기로 한다.
    <해결 방법>
    1) 우선 해야 할 일은 어떠한 오브젝트가 corrupt 되었는가를 알아내는 것이다.
    다음의 스크립트를 이용하면 알 수 있다.
    SQL>select segment_name, segment_type
    from dba_extents
    where file_id = f and
    b between block_id and block_id + blocks - 1;
    2) 만약 해당 세그먼트가 인덱스이면 drop 시키고 다시 생성하면 된다.
    3) 만약 해당 세그먼트가 테이블이면 corrupt된 블럭의 데이타는 손상된 것이다.
    4) 만약 해당 테이블이 들어있는 export 화일이 있다면 손상된 테이블을 drop
    시키고 import 받는 것이 제일 간단한 방법이다. 하지만, 만약 export 받은
    화일이 없거나 backup해 놓은 화일도 없다면 해당 테이블에 인덱스가 생성되어
    있는 경우에 한해서 다음의 방법을 사용해서 복구를 하도록 한다.
    5) 해당 테이블에 대한 인덱스가 생성되어 있다면 이를 이용해서 corrupt된
    블럭을 피해갈 수 있다. 이 방법은 다음과 같다.
    - empno, ename, deptno 를 컬럼으로 가지는 emp 테이블이 corrupt 되었다고
    가정하자. 그리고, empno 컬럼에 인덱스가 생성되어 있다고 하자. 클러스터화
    되지 않은 모든 테이블은 Unique한 rowid를 가진다.
    rowid를 varchar2/hexadecimal 형식으로 표현하려면 rowidtochar 함수를 이용
    한다.
    SQL> select rowidtochar(rowid) from emp;
    rowid는 총 18자로 block address(8자), dot(1자), row address(4자),
    dot(1자), file address(4자)로 구성되어 있다.
    SQL> select empno, rowid
    from emp
    where empno > 0
    위의 스크립트를 실행시키면 다음과 같은 결과를 얻게 된다.
    EMPNO ROWID
    100 00000003.0000.0006
    101 00000003.0001.0006
    102 00000003.0002.0006
    103 00000003.0003.0006
    500 00000004.0000.000A
    501 00000004.0001.000A
    755 0000001A.0005.000A
    756 0000001A.000C.000A
    만약 인덱스가 character 컬럼에 대한 것이었다면 위의 query 문장을 다음과 같
    이 바꿀 수 있다.
    SQL> select empno, rowid
    from emp
    where empno > '';
    예를 들어 다음과 같은 에러 메시지가 떨어졌다고 하자.
    01578, 00000, "ORACLE data block corrupted (file # 10, block # 4)
    그러면, 다음의 스크립트를 사용하여 손상된 블럭에 있는 employee 에 대한
    empno를 구할 수 있다.
    SQL> select empno from emp
    where empno > 0
    and rowidtochar(rowid) like '00000004.%.000A';
    EMPNO ROWID
    500 00000004.0000.000A
    501 00000004.0001.000A
    이제 emp 테이블과 같은 구조를 갖는 새로운 테이블을 만든다.
    SQL> create table temp
    as select * from emp
    where 1 = 2;
    그런 다음 손상된 부분을 피해서 새로운 테이블에 손상된 테이블의 데이타를
    추가한다.
    SQL> insert into temp select * from emp where empno < 500;
    SQL> insert into temp select * from emp where empno > 501;
    손상된 테이블을 drop 시키고 temp 테이블의 이름을 emp로 변경한다.
    그리고, 백업된 자료나 문서 자료를 통하여 손상된 부분에 대한 정보를 추가한다.
    6) 손상된 블럭에 여러 개의 row가 존재하고 있다면 다음의 방법을 이용한다.
    SQL> create table empnos as
    select empno from emp
    where empno > 0
    and rowidtochar(rowid) not like '00000004.%.000A';
    이 스크립트를 이용하면 손상된 block에 포함되지 않는 empno들을 알 수 있다.
    다음의 스크립트를 계속 실행시켜 복구를 한다.
    SQL> create table temp as select * from emp where 1 = 2;
    SQL> insert into temp
    select emp.empno, emp.ename, emp.deptno
    from emp, empnos
    where emp.empno > 0
    and emp.empno = empnos.empno;
    7) 만약 Data Dictionary의 table이나 index에서 손상된 block이 발생했다면
    지원을 요청해야 한다.

    Hi,
    Take a look in the following Metalink Notes:
    - DBMS_REPAIR example - Doc ID: NOTE:68013.1
    - Handling Oracle Block Corruptions in Oracle7/8/8i/9i/10g - Doc ID: NOTE:28814.
    - Prevention, Detection and Repair of Database Corruption - Doc ID: NOTE:76375.1
    Cheers,
    Francisco Munoz Alvarez
    http://www.oraclenz.com

  • Need advice on ORA-01578: ORACLE data block corrupted

    We have a development database server version- 10.2.0.3 with materialized views refresh as complete every morning. Yesterday we had a power failure and the server went down and database was shutdown unexpectedly.
    When we restarted the database after the server restarted, we found some of the datablocks got corrupted . Following were the exceptions that we saw in the alert.log.
    Errors in file /i01_01/app/oracle/product/10.2.0/db_1/admin/orcl9/bdump/orcl9_smon_7547.trc:
    ORA-01578: ORACLE data block corrupted (file # 11, block # 257712)
    ORA-01110: data file 11: '/i01_01/app/oracle/product/10.2.0/oradata/orcl9/ts_gen_data_02.dbf'
    ORACLE Instance orcl9 (pid = 8) - Error 1578 encountered while recovering transaction (9, 38) on object 54463
    I tried the following query to see the segment type.
    select owner, segment_name, segment_type from dba_extents where file_id =11 and 257712 between block_id and block_id + blocks - 1;
    OWNER
    SEGMENT_NAME
    SEGMENT_TYPE
    VISH
    INVENTORY_TXN
    TABLE
    where " INVENTORY_TXN " is a materialized view that was using the block that got corrupted. I can always recreate the MV by dropping and recreating it. Will it solve the problem???
    If not, how can I recover/repair the block.???
    Can anyone advice on this. Thanks very much in advance.

    To recover a corrupted block,the best way out is to use Blockrecover command of RMAN. So you would need RmAN backup to perform the operation.But first ypu you need to ensure that this is a persistent error or not? Is this error is coming repeatedly or just once it happened?
    About Blockrecover command,read here,
    http://download.oracle.com/docs/cd/B19306_01/backup.102/b14191/rcmrecov005.htm#BRADV157
    HTH
    Aman....

  • ORA-01578: ORACLE data block corrupted

    Dera Guru's,
    In SAP i have following error in Oracle Database
    SQL error in the database when accessing a table
    "Database error text........: "ORA-01578: ORACLE data block corrupted (file #
    38, block # 72576) ORA-01110: data file 38:
    'K:\ORACLE\D01\SAPDATA4\D01_12\D01.DATA12'"
    How can i resolve it.Please guide me...
    Thank you for your help and advice....

    Error: ORA-01578 (ORA-1578)
    Text: ORACLE data block corrupted (file # %s, block # %s)
    Cause: The data block indicated was corrupted, mostly due to software errors.
    Action: Try to restore the segment containing the block indicated. This
    may involve dropping the segment and recreating it. If there
    is a trace file, report the errors in it to your ORACLE
    representative.
    Do you already identify the object corrupted ?
    What kind of backup do you have ?
    Edited by: marcopb on Sep 10, 2012 2:47 PM

  • Oracle V11.2.0.3 ORA-01578: ORACLE data block corrupted - OBJECT = IDL_UB1$

    Hi,
    I am running into a data corruption issue.
    My database is:
    SQL> select banner from v$version;
    BANNER
    Oracle Database 11g Enterprise Edition Release 11.2.0.3.0 - Production
    PL/SQL Release 11.2.0.3.0 - Production
    CORE 11.2.0.3.0 Production
    TNS for Linux: Version 11.2.0.3.0 - Production
    NLSRTL Version 11.2.0.3.0 - Production
    The following information is written to the alert.log File
    alert.log File
    Mon Nov 07 17:24:12 2011
    Starting ORACLE instance (normal)
    LICENSE_MAX_SESSION = 0
    LICENSE_SESSIONS_WARNING = 0
    Picked latch-free SCN scheme 2
    Using LOG_ARCHIVE_DEST_1 parameter default value as USE_DB_RECOVERY_FILE_DEST
    Autotune of undo retention is turned on.
    IMODE=BR
    ILAT =27
    LICENSE_MAX_USERS = 0
    SYS auditing is disabled
    Starting up:
    Oracle Database 11g Enterprise Edition Release 11.2.0.3.0 - Production.
    ORACLE_HOME = /home/oracle/dbhome
    System name: Linux
    Node name: dbl-ora
    Release: 2.6.18-274.3.1.el5 (This is rhel5.7 or CentOs5.7)
    Version: #1 SMP Tue Sep 6 20:14:03 EDT 2011
    Machine: i686 / vm
    Mon Nov 07 19:42:14 2011
    Corrupt Block Found
    TSN = 0, TSNAME = SYSTEM
    RFN = 1, BLK = 52346, RDBA = 4246650
    OBJN = 225, OBJD = 225, OBJECT = IDL_UB1$, SUBOBJECT =
    SEGMENT OWNER = SYS, SEGMENT TYPE = Table Segment
    Errors in file /home/oracle/diag/rdbms/ora11/K/trace/K_ora_5425.trc (incident=11053):
    ORA-01578: ORACLE data block corrupted (file # 1, block # 52346)
    ORA-01110: data file 1: '/home/oracle/oradata/ora11/system01.dbf'
    Incident details in: /home/oracle/diag/rdbms/ora11/K/incident/incdir_11053/K_ora_5425_i11053.trc
    I was even able to detect the row that is generating the issue.
    In my case the obj# 33573 until 33577 are causing the issue,
    though I have no idea what sort of objects are affected.
    SQL> select * from idl_ub1$ where obj#=33572;
    OBJ# PART VERSION PIECE# LENGTH P
    33572 1 0 0 9032 F
    SQL> select * from idl_ub1$ where obj#=33573;
    ERROR:
    ORA-01578: ORACLE data block corrupted (file # 1, block # 52346)
    ORA-01110: data file 1: '/home/oracle/oradata/ora11/system01.dbf'
    no rows selected
    SQL> select * from idl_ub1$ where obj#=33577;
    ERROR:
    ORA-01578: ORACLE data block corrupted (file # 1, block # 52358)
    ORA-01110: data file 1: '/home/oracle/oradata/ora11/system01.dbf'
    no rows selected
    SQL> select * from idl_ub1$ where obj#=33578;
    OBJ# PART VERSION PIECE# LENGTH P
    33578 1 0 0 9032 F
    Any idea, how to fix this problem without recreating the whole database?
    Thanks in advance.
    wmager
    Edited by: magerxr on Nov 7, 2011 8:27 AM

    magerxr wrote:
    Thanks again for your quick advise.
    Here comes the result of dbv against my system tablespace.
    [oracle@dbl-ora ~]$ dbv FILE=/home/oracle/oradata/ora11/system01.dbf
    DBVERIFY: Release 11.2.0.3.0 - Production on Mon Nov 7 22:39:11 2011
    Copyright (c) 1982, 2011, Oracle and/or its affiliates. All rights reserved.
    DBVERIFY - Verification starting : FILE = /home/oracle/oradata/ora11/system01.dbf
    Page 52346 is influx - most likely media corrupt
    Corrupt block relative dba: 0x0040cc7a (file 1, block 52346)
    Fractured block found during dbv:
    Data in bad block:
    type: 6 format: 2 rdba: 0x0040cc7a
    last change scn: 0x0000.0010acfa seq: 0x1 flg: 0x04
    spare1: 0x0 spare2: 0x0 spare3: 0x0
    consistency value in tail: 0x00000000
    check value in block header: 0x8fda
    computed block checksum: 0xaafbselect owner, segment_type, segment_name from dba_extents
    where file_id = 1 and 52346 between block_id and block_id+blocks-1;
    >
    Page 52347 is marked corrupt
    Corrupt block relative dba: 0x0040cc7b (file 1, block 52347)
    Bad header found during dbv:
    Data in bad block:
    type: 1 format: 6 rdba: 0x0000a206
    last change scn: 0xacfa.0040cc7b seq: 0x10 flg: 0x00
    spare1: 0xfa spare2: 0xac spare3: 0x401
    consistency value in tail: 0x00000000
    check value in block header: 0x0
    block checksum disabled
    select owner, segment_type, segment_name from dba_extents
    where file_id = 1 and 52347 between block_id and block_id+blocks-1;
    Page 52361 is marked corrupt
    Corrupt block relative dba: 0x0040cc89 (file 1, block 52361)
    Bad header found during dbv:
    Data in bad block:
    type: 1 format: 6 rdba: 0x0000a206
    last change scn: 0xacfb.0040cc89 seq: 0x10 flg: 0x00
    spare1: 0xfb spare2: 0xac spare3: 0x401
    consistency value in tail: 0x32298500
    check value in block header: 0x0
    block checksum disabled
    select owner, segment_type, segment_name from dba_extents
    where file_id = 1 and 52361 between block_id and block_id+blocks-1;
    >
    >
    DBVERIFY - Verification complete
    Total Pages Examined : 122880
    Total Pages Processed (Data) : 81298
    Total Pages Failing (Data) : 0
    Total Pages Processed (Index): 22307
    Total Pages Failing (Index): 0
    Total Pages Processed (Other): 3349
    Total Pages Processed (Seg) : 1
    Total Pages Failing (Seg) : 0
    Total Pages Empty : 15910
    Total Pages Marked Corrupt : 16
    Total Pages Influx : 1
    Total Pages Encrypted : 0
    Highest block SCN : 4064615 (0.4064615)post results from 3 SQL above

  • ORA-01578 block corrupted in OLAP instances

    Hi all,
    We found in almost every instance we got OLAP Option installed this error:
    ORA-01578: bloque de datos ORACLE corrupto (archivo numero 3, bloque numero 1452)
    ORA-01110: archivo de datos 3: 'G:\ORADATA\NKDW2\CWMLITE01.DBF'
    ORA-06512: en "OLAPSYS.CWM2_OLAP_METADATA_REFRESH", linea 8
    when executing:
    cwm2_OLAP_METADATA_REFRESH.MR_REFRESH()
    The first time we thought it was due to a "real" block corruption but when started to appear
    in others instances (different physical machines) we really thought it could be a bug.
    Several disk scans on linux and windows environments showed us everything was fine but Oracle
    still persist in the block corruption.
    Any ideas? your comments are welcome
    Thanx in advanced
    aLeX

    Please confirm the corruption.
    select tablespace_name
    , segment_type
    , owner
    , segment_name
    from dba_extents
    where file_id='3'
    and '1425' between block_id and block_id + blocks -1;
    Since you're running into issues on several machines (on a procedure in a package) it's possible that a piece of the code might be corrupted in some way. Can you drop the package then recreate it? The ?/cwmlite/admin/cwm2mrrf.plb and cwm2mrrf.plb scripts recreate this bit. If it turns out you have a bad script we can wrap you a new one and send it.

  • ORA-01578: ORACLE data block corrupted (file # 1, block # 53713)

    When i tried to export data from db (Oracle 11g, 64bit on Linux)
    Im getting following error
    About to export specified users ...
    . exporting pre-schema procedural objects and actions
    EXP-00008: ORACLE error 604 encountered
    ORA-00604: error occurred at recursive SQL level 1
    ORA-01578: ORACLE data block corrupted (file # 1, block # 53713)
    ORA-01110: data file 1: '/u02/oradata/RSDB1/system01.dbf'
    EXP-00083: The previous problem occurred when calling EXFSYS.DBMS_EXPFIL_DEPASEX P.schema_info_exp
    EXP-00008: ORACLE error 604 encountered
    ORA-00604: error occurred at recursive SQL level 1
    ORA-01578: ORACLE data block corrupted (file # 1, block # 55497)
    ORA-01110: data file 1: '/u02/oradata/RSDB1/system01.dbf'
    EXP-00083: The previous problem occurred when calling SYS.DBMS_CUBE_EXP.schema_i nfo_exp
    . exporting foreign function library names for user WB_APP_MANAGER
    . exporting PUBLIC type synonyms
    EXP-00008: ORACLE error 604 encountered
    ORA-00604: error occurred at recursive SQL level 1
    ORA-01578: ORACLE data block corrupted (file # 1, block # 44638)
    ORA-01110: data file 1: '/u02/oradata/RSDB1/system01.dbf'
    EXP-00000: Export terminated unsuccessfully
    I donot understand how to solve this issue Please help me to solve this issue..
    Thanks

    891620 wrote:
    When i tried to export data from db (Oracle 11g, 64bit on Linux)
    Im getting following error
    About to export specified users ...
    . exporting pre-schema procedural objects and actions
    EXP-00008: ORACLE error 604 encountered
    ORA-00604: error occurred at recursive SQL level 1
    ORA-01578: ORACLE data block corrupted (file # 1, block # 53713)
    ORA-01110: data file 1: '/u02/oradata/RSDB1/system01.dbf'
    EXP-00083: The previous problem occurred when calling EXFSYS.DBMS_EXPFIL_DEPASEX P.schema_info_exp
    EXP-00008: ORACLE error 604 encountered
    ORA-00604: error occurred at recursive SQL level 1
    ORA-01578: ORACLE data block corrupted (file # 1, block # 55497)
    ORA-01110: data file 1: '/u02/oradata/RSDB1/system01.dbf'
    EXP-00083: The previous problem occurred when calling SYS.DBMS_CUBE_EXP.schema_i nfo_exp
    . exporting foreign function library names for user WB_APP_MANAGER
    . exporting PUBLIC type synonyms
    EXP-00008: ORACLE error 604 encountered
    ORA-00604: error occurred at recursive SQL level 1
    ORA-01578: ORACLE data block corrupted (file # 1, block # 44638)
    ORA-01110: data file 1: '/u02/oradata/RSDB1/system01.dbf'
    EXP-00000: Export terminated unsuccessfully
    I donot understand how to solve this issue Please help me to solve this issue..
    Thanksrun dbv against '/u02/oradata/RSDB1/system01.dbf'
    & post results back here
    dbv
    DBVERIFY: Release 11.2.0.1.0 - Production on Fri Oct 14 20:39:11 2011
    Copyright (c) 1982, 2009, Oracle and/or its affiliates.  All rights reserved.
    Keyword     Description                    (Default)
    FILE        File to Verify                 (NONE)
    START       Start Block                    (First Block of File)
    END         End Block                      (Last Block of File)
    BLOCKSIZE   Logical Block Size             (8192)
    LOGFILE     Output Log                     (NONE)
    FEEDBACK    Display Progress               (0)
    PARFILE     Parameter File                 (NONE)
    USERID      Username/Password              (NONE)
    SEGMENT_ID  Segment ID (tsn.relfile.block) (NONE)
    HIGH_SCN    Highest Block SCN To Verify    (NONE)
                (scn_wrap.scn_base OR scn)         

  • ORA-01578: ORACLE data block corrupted on tables in sysaux tablespace

    Dear Experts,
    From the alert log file we noticed data block corruptions on one of our datafiles. After further investigation, we realized that the corruptions were on 3 of the AWR related tables in the SYSAUX tablespace:
    1. WRH$_LIBRARYCACHE
    2. WRH$_TEMPSTATXS
    3. WRI$_ALERT_OUTSTANDING
    The bad news is that we may not have a valid rman backup to do the recovery due to the retention policy - RECOVERY WINDOW OF 2 DAYS. Since this is a development database with limited monitoring, we did not discover the corruption until 6 days later. The issue happened about 6 days ago (about Christmas time).
    So, what are our recovery options? Can someone advice? We are thinking about drop and recreate the 3 affected v$WR* tables, but not quite sure about the impact to the system if we drop and recreate the 3 objects. Did someone experience this type of recovery. If you did, what are your approaches?
    We are running oracle 10.2.0.3 version.
    I greatly appreciate your input and suggestion. Thanks!!!

    as long as you have a backup of ur database before christmas, you can use the " MAXDAYS " cmd to get ur backup working so long as you have not used delete obsolote....had a same sistuation....where i had a backup and trying to restore it ...kept saying no valid backup...after going thru some stuff...found the MAXDAYS cmd to use my backup...here is an example ...
    $ rman target /
    Recovery Manager: Release 10.2.0.2.0 - Production on Sun Apr 6 09:05:44 2008
    Copyright (c) 1982, 2005, Oracle. All rights reserved.
    connected to target database (not started)
    RMAN> SET DBID=1528894801
    executing command: SET DBID
    RMAN> startup force nomount;
    startup failed: ORA-01078: failure in processing system parameters
    LRM-00109: could not open parameter file '/u01/app/oracle/product/10.2.0/db_1/dbs/initsameera.ora'
    starting Oracle instance without parameter file for retrival of spfile
    Oracle instance started
    Total System Global Area 159383552 bytes
    Fixed Size 1259672 bytes
    Variable Size 58722152 bytes
    Database Buffers 92274688 bytes
    Redo Buffers 7127040 bytes
    RMAN> set controlfile autobackup format for device type disk to '/u99/backup/sameera/control_spfile_%F';
    executing command: SET CONTROLFILE AUTOBACKUP FORMAT
    using target database control file instead of recovery catalog
    RMAN> run
    2> {
    3> allocate channel p1 type disk;
    4> restore spfile to pfile '/u01/app/oracle/product/10.2.0/db_1/dbs/initsameera.ora' from autobackup;
    5> shutdown abort;
    6> }
    allocated channel: p1
    channel p1: sid=36 devtype=DISK
    Starting restore at 06-APR-08
    channel p1: looking for autobackup on day: 20080406
    channel p1: looking for autobackup on day: 20080405
    channel p1: looking for autobackup on day: 20080404
    channel p1: looking for autobackup on day: 20080403
    channel p1: looking for autobackup on day: 20080402
    channel p1: looking for autobackup on day: 20080401
    channel p1: looking for autobackup on day: 20080331
    channel p1: no autobackup in 7 days found
    released channel: p1
    RMAN-00571: ===========================================================
    RMAN-00569: =============== ERROR MESSAGE STACK FOLLOWS ===============
    RMAN-00571: ===========================================================
    RMAN-03002: failure of restore command at 04/06/2008 09:09:09
    RMAN-06172: no autobackup found or specified handle is not a valid copy or piece
    Solution:
    RMAN> shutdown abort;
    RMAN> EXIT;
    $ ps -ef |grep pmon
    oracle 2891 2856 0 09:05 pts/1 00:00:00 grep pmon
    oracle 7448 1 0 Apr05 ? 00:00:00 ora_pmon_primary
    $export ORACLE_SID=sameera
    $ rman target /
    Recovery Manager: Release 10.2.0.2.0 - Production on Sun Apr 6 09:05:44 2008
    Copyright (c) 1982, 2005, Oracle. All rights reserved.
    connected to target database (not started)
    RMAN> SET DBID=1528894801
    executing command: SET DBID
    RMAN> startup force nomount;
    startup failed: ORA-01078: failure in processing system parameters
    LRM-00109: could not open parameter file '/u01/app/oracle/product/10.2.0/db_1/db s/initsameera.ora'
    starting Oracle instance without parameter file for retrival of spfile
    Oracle instance started
    Total System Global Area 159383552 bytes
    Fixed Size 1259672 bytes
    Variable Size 58722152 bytes
    Database Buffers 92274688 bytes
    Redo Buffers 7127040 bytes
    RMAN> set controlfile autobackup format for device type disk to '/u99/backup/sameera/control_spfile_%F';
    executing command: SET CONTROLFILE AUTOBACKUP FORMAT
    using target database control file instead of recovery catalog
    RMAN> run
    2> {
    3> allocate channel p1 type disk;
    4> restore spfile to pfile '/u01/app/oracle/product/10.2.0/db_1/dbs/initsameera.ora' from autobackup maxdays 15;
    5> shutdown abort;
    6> }
    released channel: ORA_DISK_1
    allocated channel: p1
    channel p1: sid=36 devtype=DISK
    Starting restore at 06-APR-08
    channel p1: looking for autobackup on day: 20080406
    channel p1: looking for autobackup on day: 20080405
    channel p1: looking for autobackup on day: 20080404
    channel p1: looking for autobackup on day: 20080403
    channel p1: looking for autobackup on day: 20080402
    channel p1: looking for autobackup on day: 20080401
    channel p1: looking for autobackup on day: 20080331
    channel p1: looking for autobackup on day: 20080330
    channel p1: looking for autobackup on day: 20080329
    channel p1: looking for autobackup on day: 20080328
    channel p1: looking for autobackup on day: 20080327
    channel p1: looking for autobackup on day: 20080326
    channel p1: looking for autobackup on day: 20080325
    channel p1: looking for autobackup on day: 20080324
    channel p1: looking for autobackup on day: 20080323
    channel p1: autobackup found: /u99/backup/sameera/control_spfile_c-1528894801-20080323-00
    channel p1: SPFILE restore from autobackup complete
    Finished restore at 06-APR-08
    Oracle instance shut down
    Check to make sure if initsameera.ora exists in $ORACLE_HOME/dbs location.
    $ cd $ORACLE_HOME/dbs
    $ ls -ltr
    total 7052
    -rw-r----- 1 oracle oinstall 2560 Apr 5 13:21 spfileprimary.ora
    -rw-r----- 1 oracle oinstall 7061504 Apr 5 13:23 snapcf_primary.f
    -rw-rw---- 1 oracle oinstall 1544 Apr 5 18:42 hc_sameera.dat
    -rw-r--r-- 1 oracle oinstall 1087 Apr 6 09:12 initsameera.ora
    $ pwd
    /u01/app/oracle/product/10.2.0/db_1/dbs
    $

  • File corrupted after shutdown. (ORA 1578)

    After an accidentally shutdown my database wont start up smothly.
    If I type 'connect internal' and 'startup'
    in the svrmgrl it comes up with
    ORA-00604: error occurred at recursive SQL level 1
    ORA-01578: ORACLE data block corrupted (file #1, block #164)
    ORA-01110: data file 1:'oradata/service/system01.dbf'
    Is there anything I could do for recovery ?
    I already tried 'restore database' but without any big success.
    Help is appreciated.
    thanks,
    detlef
    PS (Please detail your suggestions as Im not an ORACLE expert.)

    After an accidentally shutdown my database wont start up smothly.
    If I type 'connect internal' and 'startup'
    in the svrmgrl it comes up with
    ORA-00604: error occurred at recursive SQL level 1
    ORA-01578: ORACLE data block corrupted (file #1, block #164)
    ORA-01110: data file 1:'oradata/service/system01.dbf'
    Is there anything I could do for recovery ?
    I already tried 'restore database' but without any big success.
    Help is appreciated.
    thanks,
    detlef
    PS (Please detail your suggestions as Im not an ORACLE expert.)

  • Oracle 11g: Block Corruption in SYSAUX File

    Hello All,
    I am facing Data corruption issue in the SYSAUX file.
    We are using Oracle 11G (Microsoft 32 bit) and our system is running in noarchivelog mode.
    Following are the errors in the alert log.
    e:\sc\sc15.2\databases\oracleconfig\diag\rdbms\enmscsdb\nm45\trace\nm45_p000_5944.trc
    Corrupt block relative dba: 0x0088a9f8 (file 2, block 567800)
    Fractured block found during buffer read
    Data in bad block:
    type: 6 format: 2 rdba: 0x0088a9f8
    last change scn: 0x0000.0b3bb7c7 seq: 0x1 flg: 0x04
    spare1: 0x0 spare2: 0x0 spare3: 0x0
    consistency value in tail: 0xc7000601
    check value in block header: 0xee6b
    computed block checksum: 0x72c6
    Reread of rdba: 0x0088a9f8 (file 2, block 567800) found same corrupted data
    Thu Jan 22 16:46:44 2009
    SMON: Restarting fast_start parallel rollback
    SMON: ignoring slave err,downgrading to serial rollback
    ORACLE Instance nm45 (pid = 12) - Error 1578 encountered while recovering transaction (9, 11) on object 458.
    Errors in file e:\sc\sc15.2\databases\oracleconfig\diag\rdbms\enmscsdb\nm45\trace\nm45_smon_6492.trc:
    ORA-01578: ORACLE data block corrupted (file # 2, block # 567800)
    ORA-01110: data file 2: 'E:\SC\SC15.2\DATABASES\ORACLECONFIG\SYSAUXNM45.ORA'
    Thu Jan 22 16:46:45 2009
    Trace dumping is performing id=[cdmp_20090122164645]
    Corrupt Block Found
    TSN = 1, TSNAME = SYSAUX
    RFN = 2, BLK = 567800, RDBA = 8956408
    OBJN = 458, OBJD = 458, OBJECT = I_WRI$_OPTSTAT_HH_OBJ_ICOL_ST, SUBOBJECT =
    SEGMENT OWNER = SYS, SEGMENT TYPE = Index Segment
    Following query indicates the corruption is in index.
    SQL> SELECT tablespace_name, segment_type, owner, segment_name FROM dba_extents
    WHERE file_id = 2 and 567800 between block_id AND block_id + blocks - 1;
    TABLESPACE_NAME SEGMENT_TYPE OWNER
    SEGMENT_NAME
    SYSAUX INDEX SYS
    I_WRI$_OPTSTAT_HH_OBJ_ICOL_ST
    ==============
    DBverify output:
    ==============
    E:\SC\SC15.2\Databases\OracleConfig>dbv file=SYSAUXNM45.ORA blocksize=8192
    DBVERIFY: Release 11.1.0.7.0 - Production on Thu Jan 22 16:59:11 2009
    Copyright (c) 1982, 2007, Oracle. All rights reserved.
    DBVERIFY - Verification starting : FILE = E:\SC\SC15.2\Databases\OracleConfig/SY
    SAUXNM45.ORA
    DBV-00200: Block, DBA 8956312, already marked corrupt
    Page 567800 is influx - most likely media corrupt
    Corrupt block relative dba: 0x0088a9f8 (file 2, block 567800)
    Fractured block found during dbv:
    Data in bad block:
    type: 6 format: 2 rdba: 0x0088a9f8
    last change scn: 0x0000.0b3bb7c7 seq: 0x1 flg: 0x04
    spare1: 0x0 spare2: 0x0 spare3: 0x0
    consistency value in tail: 0xc7000601
    check value in block header: 0xee6b
    computed block checksum: 0x72c6
    DBVERIFY - Verification complete
    Total Pages Examined : 1623864
    Total Pages Processed (Data) : 540984
    Total Pages Failing (Data) : 0
    Total Pages Processed (Index): 964944
    Total Pages Failing (Index): 0
    Total Pages Processed (Other): 17849
    Total Pages Processed (Seg) : 0
    Total Pages Failing (Seg) : 0
    Total Pages Empty : 100086
    Total Pages Marked Corrupt : 2
    Total Pages Influx : 1
    Total Pages Encrypted : 0
    Highest block SCN : 190789648 (0.190789648)
    SQL> select * from v$database_block_corruption;
    FILE# BLOCK# BLOCKS CORRUPTION_CHANGE# CORRUPTIO
    2 567800 1 0 FRACTURED
    2 567704 1 0 FRACTURED
    How to resolve this issue.
    Thanks
    With Regards
    Hemant Joshi.

    Drop and re-creating the index would be better than re-building the index.
    Check the metalink note: Identify the corruption extension using RMAN/DBV/ANALYZE etc - 836658.1
    Note 28814.1 - Handling Oracle Block Corruptions in Oracle7/8/8i/9i/10g
    Note 472231.1 - How to identify all the Corrupted Objects in the Database reported by RMAN
    ORA-1578 Main Reference Index for Solutions -830997.1

  • System datafile block corruption - no backups and database in NOARCHIVELOG mode

    Dear All,
    Database version - oracle 11.1 Enterprise
    OS - RHELinux 5.8
    What are the options of recovering from physical block corruption when there are no backup available to perform block media recovery?
    V$DATABASE_BLOCK_CORRUPTION reports two segments corrupted (please see attached image for details).
    1 table in system datafile - sys_fba_barrierscn
    1 index - (applicaiton index)
    What are my options?
    I know there is a possibility that the database will not restart after a shutdown due to corruption in system tablespace.
    Database is in noarchivelog mode. So online backups are not possible and there aren't any full backups either.
    I am thinking of below,
    1. Run dbms_repair with fix_block_corruption. - Still database startup might fail?
    2. Shutdown the database and take offline full backup with RMAN MAXCorrupt option.
    Appreciate your suggestions and advises.
    Thanks
    Stefan

    Thanks Sybrand,
    Agree with your first two suggestions .
    Also scheduled a expdp job tonight. (Only backup like thing they had was a expdp cron, but until today all the large tables were failing due to small undo_retention).
    Yes. Flashback is not used. So hopefully it will not affect the a database restart i guess?
    Related to dbms_repair, I was referring to - http://askdba.org/weblog/2010/08/physical-corruption-ora-1578-part-3/.
    Where DBMS_REPAIR.FIX_CORRUPT_BLOCKS and DBMS_REPAIR.SKIP_CORRUPT_BLOCKS used.
    Which i think will not use any redo.
    Thanks
    Stefan

  • Oracle 11g - How to repair block corruption(on free space) in datafile

    Hi,
    I have a tablesopace with 3 datafiles, out of which one datafile has corrupted block. But no objects or data is affected as the corrupted block os in free space. This was shown in the alert logs.
    Please see below the details:
    Wed Apr 06 15:30:04 2011
    SMON: Restarting fast_start parallel rollback
    SMON: ignoring slave err,downgrading to serial rollback
    ORACLE Instance geooap (pid = 12) - Error 1578 encountered while recovering transaction (10, 6) on object 149755.
    Errors in file f:\oracle11g\diag\rdbms\geooap\geooap\trace\geooap_smon_5540.trc:
    ORA-01578: ORACLE data block corrupted (file # 7, block # 54053)
    ORA-01110: data file 7: 'F:\ORACLE11G\ORADATA\GEOOAP\ORDER_DATA_01.DBF'
    GEOAP:
    Fri Apr 01 14:57:48 2011
    Errors in file f:\oracle11g\diag\rdbms\geop\geop\trace\geop_arc1_2156.trc:
    ORA-00235: control file read without a lock inconsistent due to concurrent update
    Fri Apr 01 14:57:58 2011
    ================================================================
    The corruption is being reported in a free space block of the ORDER_DATA_01.DBF.
    I’ve checked all the tables (and indexes) in this tablespace and none report corruption.
    =====================================================Is there any action I need to take to remove corruption at this point?It is not affected any operation on the database yet.
    What is the best way to do get rid of the corrupt block, without dropping and rebuillding the full tablespace(which is around 6 GB -total of 3 datafiles)
    Thanks a lot

    Can RMAN recover the datablock from this cold backup(which is a week old, the data file was not corrupted then) ?Please note that to do the recovery, you would need the backup and the archivelog files since the backup. Think about what you are asking to do. Its a single block whose recovery you are asking from a week old backup which obviously would be on an much older SCN compared to the rest of the database. How would you make that block consistent with the rest of the datafile. If you don't have archivelog in that db whose block is corrupted, you may forget that block and any data that it might ever had. Also, please read the documentation about the block recovery which explains the requirements very clearly,
    http://download.oracle.com/docs/cd/E11882_01/backup.112/e10642/rcmblock.htm#BRADV89784
    From the above link, 1st point,
    The target database must run in ARCHIVELOG mode and be open or mounted with a current control file.HTH
    Aman....

Maybe you are looking for

  • How do I make my iMac g4 wireless?

    I have an iMac g4 with 10.4 running. How can I make it wireless? If I can where is a good place to buy the part. I know apple does not make it anymore. Thanks for any help in advance.

  • Can't open attachments in mail

    Mac OSX 10.9.4 Can't open any attachments in mail by double clicking. Was able to do previously. Now need to click on attachments, download, then open from application (word, acrobat, excel, etc). Even when attachment is in downloads folder, double c

  • UPPER() in external table definition

    Hi, Is it possible to apply the UPPER() function to character columns within an external table definition using the ORACLE_LOADER driver? I've tried myself and have looked through the documentation, and it seems like this is not possible ... just wan

  • Problems with photoshop elelements 11 - can't handle images larger than appr 15-40 MB (jpeg)

    Someone who knows the limitastions of PE 11 about possible size of images (jpeg)?

  • Hai gurus...

    hello guru's... in select single and select upto one row which one is best performance wise and which one will take less time and also which one will give us the correct result and also can i know how these is related to the BUFFERING concept....