ORA-1578 ORA-354 ORA-600 [3020] Misplaced blocks by Symantec / Veritas after adding LUN

(Doc ID 1323532.1)

Last updated on DECEMBER 26, 2012

Applies to:

Oracle Server - Enterprise Edition - Version 10.2.0.4 and later
Information in this document applies to any platform.

Symptoms

ORA-1578 / ORA-354 or ORA-600 [3020] during recovery due to misplaced blocks after adding a LUN by Symantec / Veritas.  A misplaced block means that a block content was written in a different location (different file or different block) overwriting the original block.  The corruption is reported in the alert log and by dbverify with message "Bad header found":

Corrupt block relative dba: 0x01434cd4 (file 5, block 216276)
Bad header found during buffer read
Data in bad block:
 type: 6 format: 2 rdba: 0x028364bc
 last change scn: 0x0606.4c561c33 seq: 0x1 flg: 0x04
 spare1: 0x0 spare2: 0x0 spare3: 0x0
 consistency value in tail: 0x1c330601
 check value in block header: 0x6f87
computed block checksum: 0x0
Reread of rdba: 0x01434cd4 (file 5, block 216276) found same corrupted data



The block is healthy (computed block checksum is 0x0) but the content belongs to a different block.  In this example rdba: 0x01434cd4 (rfile#=5, block#=216276) was overwritten with the content of rdba: 0x028364bc) = (rfile#=10 block#=222396)

If disk mirroring is in place a reread of the same block may return a good version of the block because the other side of the mirror may not be corrupt.  In this case subsequent dbverify runs may also not report the block as corrupt.  Example from reading a good version from disk printed in the alert log:

Corrupt block relative dba: 0x01434daf (file 5, block 216495)
Bad header found during user buffer read
Data in bad block:
 type: 6 format: 2 rdba: 0x02836397
 last change scn: 0x0606.4c562f09 seq: 0x1 flg: 0x04
 spare1: 0x0 spare2: 0x0 spare3: 0x0
 consistency value in tail: 0x2f090601
 check value in block header: 0x7480
 computed block checksum: 0x0
Reread of rdba: 0x01434daf (file 5, block 216495) found valid data


This may affect any other database files like redo log files/archived logs producing ORA-354.  For an archived log the error reported in the alert is still the same:

Corrupt block seq: 38638 blocknum=374784.
Bad header found during backing up archived log
Data in bad block - seq:0. bno:0. time:0
beg:0 cks:0
calculated check value: 0
Reread of seq=38638, blocknum=374784,file=/data/pwbbr02/archives/PWBBR02_0001_0000038638_0710761678.arc


This could also cause a missing write in the original block were the IO was intended to go to in the first place (rfile#=10 in our example above) for which it may be needed to restore the original file and apply media recovery; although this block is not corrupt but has stale/old data.  The recovery of this block may not fix the stale version if the block is modified after the miss write happened producing ORA-600 [3020]

Changes

A new LUN has been added with PERSISTENCE set to NO while the Oracle database is OPEN.

Cause

Sign In with your My Oracle Support account

Don't have a My Oracle Support account? Click to get started

My Oracle Support provides customers with access to over a
Million Knowledge Articles and hundreds of Community platforms