Re: Bug 3798351 Importing a pre-existing table may DROP the table on an error
Date: Wed, 23 Jun 2010 07:28:13 +0200
Yes, I was surprised as well. I swear an import never dropped an table in my life and I was so 100% sure that it isnt possible but looks like everything is possible in Oracle but of course they can get sued.
Just because of this bug 184.108.40.206 and 220.127.116.11 must be prohibited IMHO.
-- LSC On Wed, Jun 23, 2010 at 5:10 AM, Robert Freeman <robertgfreeman_at_yahoo.com>wrote:Received on Wed Jun 23 2010 - 00:28:13 CDT
> Am I getting this right... an import failed, raised an error, and then as a
> result of the error the import program dropped a table? Do you have a bug
> number for this?
> Robert G. Freeman
> Master Principal Consultant, Oracle Corporation, Oracle ACE
> Author of various books on RMAN, New Features and this shorter signature
> Blog: http://robertgfreeman.blogspot.com
> *From:* LS Cheng <exriscer_at_gmail.com>
> *To:* Oracle Mailinglist <oracle-l_at_freelists.org>
> *Sent:* Tue, June 22, 2010 3:17:04 PM
> *Subject:* Bug 3798351 Importing a pre-existing table may DROP the table
> on an error
> I have a customer hit this nasty bug this morning.
> Really really nasty which affects 18.104.22.168 and 22.214.171.124. It turns out
> customer was trying to import a few thousands rows to the applicacion main
> table (partitioned), composed by 800 million rows, the first import gave an
> error complaining buffer size too small and that triggered a DROP TABLE
> which was lighting fast by the way, dropping the table and all partitions
> took around 15 seconds, we suspect that it didnt even try to drop the table
> physically, just updated data dictionary.
> So what a disgrace my customer said, Oracle Software actually deleted his
> most important database table and the doubt is what is Oracle responsability
> in this situation? The system has been stopped 10 hours already due to this
> bug (database is around 2TB and requires complete restore).