你在这里

You are here:

首页

博客

PDSERVICE的博客

Offer for recovering oracle data could not open database due to ora-01251:unknown file header version on one datafile

Posted by PDSERVICE on 10月 23, 2018 In

Question:

we have problem with corrupted file in ORACLE data base.

Please answer us if you could save the data, and senm us offer for aplication to restore corrupted data.

Error is below:

could not open database due to ora-01251:unknown file header version on one datafile

dbverify - verification starting: FILE = xx.dbf

page 1 is influx - most likely media corrupt

corrupt block relative dba: 0x04000001 (file 16, block 1)

fractured block found during dbv:

data in bad block:

type:11 format:2 rdba:0x04000001

last change scn:0x0000.00000000 seq:0x1 flg:0x04

as per issue datafile 16 is corrupt.

if db is noarchivelog and bo backup we can not recover the file.

Regarding DUL i already updated, we will need to consult with your management and determine if the data is important enough to contact Oracle onsite consult.

Answer:

we can provide dul recovery service , can you pls try PRM-DUL first?

download link: http://zcdn.parnassusdata.com/DUL5108.zip

Question:

I've alredy tested the free edition of DUL4108.zip.

It seems that it managed to unload table data from the datafile with the corrupted datafile header.

The problem is I had to test with small scratch tables whose segments do not span multiple datafiles because of the 10.000 rows limit.

So we still cannot be 100% sure whether it will work with a table that spans multiple datafiles and that has blocks belonging to the datafile with

corrupted file header. Is there a way to test-prove it ?

The other issue I encountered is with varchar datatype. The database is created with EE8MSWIN1250 characterset, but the diacritic characters are

scrambled in the sqlloader file created with the PRM-DUL "unload and save as" option. From what I've seen, PRM-DUL correctly recognized the

database characterset when scanning the content of the system01.dbf.

Answer:

1、 Is there a way to test-prove it ? no , but if it can recover 10000 rows , it means it can recover more.

2、 but the diacritic characters are

scrambled in the sqlloader file created with the PRM-DUL "unload and save as" option.

pls use the data bridge mode , and make sure the target database charset is EE8MSWIN1250 , then there won't be charset problem.

the prm-dul enterprise edition costs 1500 USD per database . I will send you a paypal payment link , after payment you will receive the key.

We provides recover service , thats means you no longer need to do it yourself . The price is up to the Database size .

Are you using the dictionary mode or non-dictionary mode? if the non-dictionary mode , you need the choose the correct charset yourself.