RE: Another Database Down

Laura,

I fight this nearly every damn day. We constantly refresh several SAP systems from production copies via EMC BCV's and TimeFinder. About half of the time, I get the same error, where recovery was successful, but open resetlogs would fail. It turns out, in my case, that no amount of recovery will every get this database open (even with resetlogs) with this backup set. Apparently there is an issue with a piece of the EMC software that causes problems when mirrors are split during a hot backup. To make a long story short, there is an event you can set in via a server manager session that will dump all datafile headers. What you look for is what Oracle calls a "fuzzy SCN". That is, an SCN hex value that evaluates to a nonsensical  number. In English, the SCN for the header should have some hex value that corresponds to a number. If it is set to something like FFxFFFFFFFF, then you have a "fuzzy SCN" which means your backup is bad and will never work. I don't know if this is your exact problem or not, but since you're getting the same symptoms I was, I thought I would throw it out there. You can read more on MetaLink by searching for "fuzzy SCN". By the way, as soon as our RS/6000 SA upgraded our version of PowerPath, the bug went away, we have consistent mirror splits, and all is well on the SAP backup/restore front...

Sorry if I threw you way off on a tangent on this one.

Jim

Jim Hawkins
Lead SAPR/3 Oracle Database Administrator MEMC Electronic Materials, Inc.
600 Pearl Drive
St. Louis, MO 633376
9636) 474-7832
jhawkins_at_memc.com (work)
jhawkins_at_primary.net (home)

> <snip>
> Just to put you a little more in the picture I have recovered a couple of
> DBs now without archived redo log files etc where all hope had been lost
and
> resumes were being updated and got them back by fudging a couple of
issues.
> (Oracle would probably throw their hands up in horror and refuse to
support
> what I did, but the database opened, I didn't lose any data (this I
believe
> was down to good fortune rather than skill) and everyone was happy.
> (snip>
> Lee
> <snip>
>
> Lee, I am just glad someone has responded and wants to help!
>
> I am not sure how much I told you about my situation, so bear with me
while
> I recite it to you. Our OS is Windows NT. We had 2 disks to crash. Two
of
> my databases are ok because I had archive logs for one and a cold backup
for
> the other. However, I have one that I have not done anything too as far
as
> backup, other than a physical backup of the database datafiles, redo logs,
> etc. The Sysadm had a full backup of the previous Friday and a
differential
> backup from the following Thursday which he loaded back to the server when
> the new disks were installed.
>
> I tried to recover the database and when it asked for the logs I entered
> cancel. It said that recovery was successful but gave me a warning that
if
> I tried to open the database with the Resetlogs option it would tell me
that
> my system datafile was too incomplete (or something to that affect).
>
> Yesterday (Thursday) I did call Oracle and this is what this first guy had
> me do: 1) Startup Mount 2) Select from v$datafile to check the status
of
> my datafiles. They were all online, which per Support, that was not good.
> 3) Alter database datafile ??? offline. This command gave me the error
> 'offline immediate disallowed unless media recovery enabled (ORA 01145).
>
> So then I was referred to another Support person. He had me shutdown the
> database, startup mount, recover database. Of course the recover asked
for
> a log and Support had me input each of my five redo logs, but the log with
> the info needed had already been overwritten so the info wasn't there and
I
> didn't get any further.
>
> Next Support wanted me to restore my files from the Friday backup only.
> This is where I have stopped because now I am waiting on my Sysadm to come
> in and restore them for me. I tried to call the Support guy back and let
> him know the status, but got a message stating that because I had only
> 'basic or bronze' support that I could only speak to someone between
certain
> hours, and then was cut off. I had gone through Metalink though so I just
> updated the tar.
>
> If you have any ideas for me I would appreciate any and all.
>
> Thanks,
> Laura
>
>

-- 
Jim Hawkins
Lead SAPR/3 Oracle Database Administrator
MEMC Electronic Materials, Inc.
600 Pearl Drive
St. Louis, MO  633376
9636) 474-7832
jhawkins_at_memc.com (work)
jhawkins_at_primary.net (home)

-- 
Please see the official ORACLE-L FAQ: http://www.orafaq.com
-- 
Author: Jim Hawkins
  INET: jhawkins_at_primary.net

Fat City Network Services    -- (858) 538-5051  FAX: (858) 538-5051
San Diego, California        -- Public Internet access / Mailing Lists
--------------------------------------------------------------------
To REMOVE yourself from this mailing list, send an E-Mail message
to: ListGuru_at_fatcity.com (note EXACT spelling of 'ListGuru') and in
the message BODY, include a line containing: UNSUB ORACLE-L
(or the name of mailing list you want to be removed from).  You may
also send the HELP command for other information (like subscribing).