Howard J. Rogers wrote:
> "John Oakes" <john_at_networkproductions.net> wrote in message
> news:zqYvb.27269$M31.625393_at_twister.tampabay.rr.com...
>
>>"Sybrand Bakker" <gooiditweg_at_sybrandb.nospam.demon.nl> wrote in message
>>news:bth0svoa2a05bkhtpq40qbr13ctitmu1tv_at_4ax.com...
>>
>>>On Sun, 23 Nov 2003 04:24:02 GMT, "John Oakes"
>>><john_at_networkproductions.net> wrote:
>>>
>>>
>>>>Ok, I was able to find the corrupted blocks succesfully. But now, I
>
> want
>
>>to
>>
>>>>use DBMS_REPAIR.FIX_CORRUPT_BLOCKS to mark them as bad, and I am
>
> getting:
>
>>>>DECLARE num_fix INT;
>>>>*
>>>>ERROR at line 1:
>>>>ORA-00600: internal error code, arguments: [kdrp_chkblk_1], [0], [],
>
> [],
>
>>[],
>>
>>>>[], [], []
>>>>ORA-06512: at "SYS.DBMS_REPAIR", line 287
>>>>ORA-06512: at line 4
>>>>
>>>>Any ideas? Thanks!
>>>
>>>
>>>ora-600 ----------> CALL SUPPORT
>>
>>Thanks, I have located the exact bug that causes this (found it in
>>Metalink), it is number 2021123. According to the description it is fixed
>>in product version 10.0. I'm not familiar with the Oracle versions, if I
>>upgrade from 8.1.7.0 to 8.1.7.4 is this version 10? What does version 10
>>relate to? Thanks!
>
>
> No, 8i is version 8.
> 9i is version 9.
> 10g is version 10.
>
> And 10g has been out about a month and a half, so is still extremely new
> (and likely has bugs all of its own to contend with).
>
> I would have thought that it was a bit extreme to upgrade two versions of
> the database merely to fix a bug in dbms_repair. Especially when (reading
> the bug report) the fix is likely to be nothing more than to signal a proper
> ORA- error rather than throw an ORA-600. The issue is that your block which
> is corrupt is not a data block nor an index block (it could therefore be a
> rollback segment header block, or a tablespace bitmap block, etc etc).
> DBMS_REPAIR is not going to fix such blocks up, whatever version you have.
> So the issue is simply how dbms_repair reports the problem.
>
> Regards
> HJR
What is out is 10.1.0.1 also known as Beta 2. Out ... is not out as in
out for the public. The OP will need to find another work around as a
delay of several months will likely be unacceptable to management.
--
Daniel Morgan
http://www.outreach.washington.edu/ext/certificates/oad/oad_crs.asp
http://www.outreach.washington.edu/ext/certificates/aoa/aoa_crs.asp
damorgan_at_x.washington.edu
(replace 'x' with a 'u' to reply)
Received on Sun Nov 23 2003 - 13:14:45 CST
