Re: 2 questions, deleting a tablespace on a damaged instance and tablespace

In <3f781b2b$0$28898$afc38c87_at_news.optusnet.com.au> "Howard J. Rogers" <howardjr2000_at_yahoo.com.au> writes:

>Richard Foote wrote:
 

>>
>> "Howard J. Rogers" <howardjr2000_at_yahoo.com.au> wrote in message
>> news:3f774ec9$0$32058$afc38c87_at_news.optusnet.com.au...
>>> Stan Brown wrote:
>>>
>>> > First, let me apologize fro starting a new thread on this, but I seem
>>> > to have stopped getting any replies to the old one :-(
>>>
>>>
>>> I tried replying to your emails, but every time I got a 'yahoo is not a
>>> trusted emailing source, and your email has thus been blocked' reply.
>>>
>>> >
>>> > I've got a V7 instance that suffered a hardware failure, while doing a
 hot
>>> > backup. The disk that failed was the disk contain the archivelogs. In
>>> > addition I discovered that my backup methodology had a 2G limit that I
 was
>>> > unaware of, so my backups of the .dbf files are useless. In addition, I
>>> > tried doing media recovery, when perhaps I should not have.
>>> >
>>> > In any case, here are the questions.
>>> >
>>> > 1. I can get as far a "startup mount". at that point what do I need to
 do
>>> > to delete a tablespace (contained in a single .dbf file) from the
>>> > instance?
>>>
>>> Alter database datafile 7 offline drop;
>>>
>>
>> Hi Howard,
>>
>> Slight correction. The drop clause only makes sense with a noarchivelog
>> database as by setting the file offline in this way effectively means the
>> data file is doomed.
 

>Correct. That's what Stan said in earlier posts on this subject. The data
>file in question is one that contains indexes only. And they can be, er,
>rebuilt!!
 

>>The drop is there to let Oracle know you "really"
>> want to drop it.
 

>Precisely.
 

>>If the database is in archivelog mode, forget the drop
>> clause and simply drop the tablespace after opening the database (if
>> dropping the tablespace is the intention).
 

>True, I suppose. But since the one form of the command does perfectly good
>duty in both cases (ie, archivelog and noarchivelog), I prefer to stick to
>the one form of the command.
 

>Poor Stan has been throwing this one round the group for about a week now,
>and now my reply gets commented on, so he'll probably wonder whether my
>original suggestion will work or not. Maybe you'd care to confirm for him
>that, whether or not my suggestion was purist-ly correct, it would have the
>desired effect??

It's not really a week, the actual failure was at ~ 23:00 last Wednesday night.

O _greatly_ appreciate the help I have been getting from you on this. I suspect if it wasn't for the mess email has become these days, we would already have this problem solved.

I apologize for my lack of knowledge in this area, and greatly appreciate any assistance that anyone cares to provide.

Having made some mistakes early in the attempted recovery process, I'm Leary to try things expirimentaly on this now.

-- 
"They that would give up essential liberty for temporary safety deserve
neither liberty nor safety."
						-- Benjamin Franklin