Re: How to apply archive logs on an cold backup

From: Frank van Bortel <fvanbortel_at_netscape.net>
Date: Fri, 03 Sep 2004 11:23:17 +0200
Message-ID: <ch9cp7$5iv$1@news5.tilbu1.nb.home.nl>

Howard J. Rogers wrote:
> Frank van Bortel wrote:
>
>

>>Gents,
>>
>>it seems to me there is some misunderstanding as to
>>what the OP is trying to accomplish; 

>
>
> No, there isn't. Not this side of the equator at least.

OK - hey, listen, I'm sorry - OK? Cool down; jeez, you *are* itchy these days. Why do you think: * I did not just address it to you, but started off with 'gents' - let me beat you to it: not implying you're no gentleman.
* I used the words 'it seems to me' - there's aspects of uncertainty and *personal* thought in that.

>
>

>>as I understand 
>>it the OP wants:
>>- some sort of PITR (as close to "now" as possible)

>
>
> Excuse me? Here is the original post. In full.
>
> "i have the database with archive log and one cold backup.
> how to apply archive log in cold backup?"
>
> Now how in Heaven's name you can intuit that somewhere in that sentence is
> ANYTHING to do with a PITR, I haven't the faintest idea. But if you've got
> a fresh stock of crystal balls, be my guest.

No i don't, I just try to imagine what the OP's situation is. And that's the closest I get, reading his detailed, elaborated (not!) post.
>
> Just don't start saying my answer is wrong when what you've actually done is
> completely change the question!
>

This (part of this) thread was not soley intended for you (I just had to reply somewhere...) to start with, so do not feel spoken to when not.
And don't put words in my mouth, either. There is not one sentence in the -admittingly crappy- english I wrote, that could let you believe I was saying your answer was wrong - where did you read that?!?
>

>>- without having a proper hot backup scenario

>
>
> He has a proper backup scenario. That's all one ever needs to recover a
> database. Hot or cold, makes no difference to someone in archivelog mode.
> Why is this message not getting through?
>
>

>>AFAICT, the OP wants to recover (PITR), but only happens
>>to have a cold backup, and all online, and off-line
>>redo log files.
>>Now, you can restore your cold*) backup, and everything
>>is honkey-dorey;

>
>
> It makes not the slightest difference whether he's got a cold backup, a hot
> backup, or wants to do a complete or incomplete recovery.
>
> The answer is the same in all cases: restore what's wrong, and start
> applying redo to it.
>
> If, as you say, he wants to do a PITR, how do you do that? You restore all
> data files *alone*, and issue a recover database until command. No backup
> controlfile clause.
>
> Of course, if you're saying he wants to do a point in time recovery AND has
> lost his control files, then naturally he will restore his binary control
> files as well as his datafiles, and equally naturally he will then have to
> issue the 'using backup controlfile' command during recovery.
>
> The commands he issues, and whether a particular clause is required or not,
> depends entirely and utterly on you defining, PRECISELY, what he has lost,
> how he intends to replace them and what he hopes to achieve at the end of
> it. It does *not* depend on whether he has a hot or a cold backup.
>
> And 'using backup controlfile' is used in ONE of those scenarios, and is not
> the SOLE answer as Bob's reply implied. And it is not even the appropriate
> answer if all he wants to do is a point in time recovery due to, say, a
> user dropping a table.
>
>

>>however, you lost 2 days of data. 
>>Where trying to recover those two days, two scenario's
>>are possible (at least):
>>The OP does still have the original controlfile(s)
>>and online redolog files, or he doesn't.
>>
>>As he has not stated that, nor reacted to this thread,
>>I guess the rest is academically.
>>
>>Howard - why would this scenario never arise? 

>
>
> What scenario? Using backup controlfile?

The scenario I tried to depict in the previous thread.

>
> When did I ever say it never would arise??

Just reread your contributions to this thread

>
> It's bad enough just completely re-writing the original post to make your
> point, but please don't put statements like that in my mouth either.

I'm not making a point. I'm asking questions, ventilating opinions. Just read on...

>
> What I said to Bob was that it was stupid to tell someone to start issuing
> 'recover database ... using backup controlfile' commands when you have no
> idea what recovery he actually wants to perform, and you have no idea what
> failure situation you're dealing with. And to mention ONLY that option is
> stupider, because it carries risks.
>
> The correct reply to the original question is: you apply redo to your backup
> as anyone else does: restore the bits of your backup that need restoring,
> and recover it as appropriate. And the temperature of the backup concerned
> has NO effect on that answer.
>
> You can't get more specific an answer than that because the original post,
> as actually written, and not as fondly imagined to be by other contributors
> to this thread, allows no greater degree of specificity.
>
>

>>Your 
>>hot backups just aren't usable (tapes: rust on wheels),
>>that's all it takes, I'd say.
>>And not verifying your tapes is a non-proper backup
>>scenario in my book.

>
>
> None of that has any bearing, I'm sorry to say, on anything the original
> poster asked or the reply I gave.
>
> HJR
>
>

No, but with the scenario I tried to depict; I'm sorry for my obvious lack of understanding of the English language; I'll try to get my thoughts accross in Dutch next time.

-- 

Regards,
Frank van Bortel

Received on Fri Sep 03 2004 - 04:23:17 CDT