Email: [email protected] 7 x 24 online support!
ora600 kcbzpbuf_1 crashed and can’t startup again due to block corruption
I have a customer who is using 10.2.0.5 ,and currently they faced a storage issue which caused the data block corruption , and we have already tried
10153,10515 10061 and _corrupted_rollback_segments and _allow_resetlogs_corruption to get this started, previously it can get it started , however
	now it crashed again , and it crashed due to ORAkcbzpbuf_
	1.
	AS of now we just want to get this database open ,any other experience is appreciated.
Alert log file:
	Hex dump of (file 3, block 155049) in trace file /oracle/app/oracle/product/10gRDB/rdbms/log/fmsdb3_dbw6_8325.trc
	Corrupt block relative dba: 0x00c25da9 (file 3, block 155049)
	Bad header found during preparing block for write
Data in bad block:
	type: 60 format: 6 rdba: 0x00c25da9
	last change scn: 0x0722.8006df22 seq: 0x1 flg: 0x00
	spare1: 0x65 spare2: 0x81 spare3: 0x0
	consistency value in tail: 0xdf223c01
	check value in block header: 0x0 >>
	is it expected?
	block checksum disabled
	Fri May 17 21:09:19 SAT 2013
	Errors in file /oracle/app/oracle/product/10gRDB/rdbms/log/fmsdb3_dbw6_8325.trc:
	ORA00600:
	Message 600 not found; No message file for product=RDBMS, facility=ORA; arguments: [kcbzpbuf_1] [4] [1]
	Dbw Trace file :
	BH (c00000035efc23d8) file#: 3 rdba: 0x00c25da9 (3/155049) class: 1 ba: c00000035e8f8000
	set: 31 blksize: 8192 bsi: 0 setflg:
	2 pwbcnt: 0
	dbwrid: 6 obj: 1
	objn: 1
	tsn: 2 afn: 3
	hash: [c0000005d444b590,c0000005d444b590] lrureq:
	[c0000005d398e558,c0000005d398e558]
	lruflags:
	on_auxiliary_list
	ckptq: [c0000005cf57a500,c0000005cf57a500] fileq: [NULL] objq: [NULL]
Answer:
This error means the database writer has prepared a block to be written from memory (buffercache) to disk, but before
	doing the actual write the block is checked once more to see if everything is OK. However, this final check finds the
	block has become corrupted.
	It means the root cause of the corruption I still there. It is most likely a faulty disk or a diskdriver.
	It must be rectified
	first before you attempt to open the database.
	In such cases the database writer is designed to report this error and crash the instance to preserv the integrity of the
	data.
	In case you do not have a backup of the database and need to extract the data, I suggest you do it using the DUL
	utility. Just observe DUL cost extra for the customer.
pls check DUL here :http://www.parnassusdata.com/en/emergency-services
