Re: Fssnap and Oracle 9i

From: Holger Baer <holger.baer_at_science-computing.de>
Date: Wed, 20 Oct 2004 12:39:16 +0200
Message-ID: <cl5f8l$58j$1@news.BelWue.DE>

Howard J. Rogers wrote:
> Holger Baer wrote:
>

[..]

>

>>Now while your answer was precise and correct, 

>
>
> Thank you. I thought so too. Joel has, of course, said that it was "below
> par", sub-standard and incorrect.
>
>

>>it left me too 
>>wondering, why you didn't mention what Oracle offers to make
>>hot backups working. It's just not you to leave that out.

>
>
> Because the OP didn't ask about them. The context of his question was, I
> have this tool. Will it do the job safely? Now, I could have mentioned
> 'yes, if you write a begin backup script'. To my mind, and in my way of
> looking at things, that is effectively a "no, not unless you do lots of
> other hoop-jumping".

I've known you to give people not necessarily what they asked for, but instead told them the right questions, which in my opinion should include Oracle hot backups.

That this includes 'lots of other hoop-jumping' is something you often take the time to painstakingly draw the big picture (I know I learned a lot from such posts) and don't mind at all.

Some time ago there was a discussion on how to recover through resetlogs, and I believe you wrote something a long the lines 'this is just a demonstration of, look how clever I can be by disobeying Oracles recommendations which are: after open resetlogs shutdown your instance and take a full cold backup'. IIRC it took you about 2 or 3 sentences to summarize what you need for recovery through resetlogs. Maybe the discussion was on something else and you just used this as an example, my memories ain't what they should be at my age ;-)

[...]
> It's a difference in perspective, I suppose, and we are *all* entitled to
> have them without one-liner attempts to make one party sound foolish or
> incompetent for seeing it that way. In my estimation, and by my reading of
> the original post, the guy wanted to know whether his backup tool would
> work, as is, without additional scripting. And hence the answer I gave him.

Yes, but somehow for the innocent bystander it looked like if fssnap is not oracle aware, then there won't be any way to use it in a backup scenario other than shutting down the instance. But I see what you mean.

[...]
>

>>So Joel stepped in and gently asked about hot backups. I'm neither
>>natural english nor had an education in languages, so I might
>>missed some fine nuances of his answer, but I was really surprised
>>how you reacted. And the rest of the thread really should not have
>>happend.

>
>
> Your definition of "gently" is not mine, which is fine.

Ok, not so gentle on reflection.

>
> All Joel went on to say was "Yeah, I think he should use RMAN too". But it
> was what came before that which shouldn't have happened.
>
> Don't you think, if all Joel wanted to do was 'gently' question why I hadn't
> mentioned hot backups he could have posted 'Howard, just curious: why
> didn't you mention hot backups?'.

That's the way I understood his reply in the first place, but what ever.
>
> I accept that it is a subtle nuance between that and "Isn't that what hot
> backups are for?" And "mostly true and correct, but why wouldn't hot
> backups [do]..."
>
> But in a world where there is a profound difference between "keep" and "pin"
> that is not a negligible difference.
>
> I found Joel's comments personally slighting, which is one thing and I could
> live with it. But I also found them unhelpful to the original poster. He
> didn't explain hot backups to the OP, just asked me (snidely, I thought)
> why they wouldn't work. In the process, he 'tainted' my answer, making it
> out to be incomplete and wrong, though he ended up agreeing with its
> ultimate conclusion! That is likely just to be confusing for the OP. And
> it's that which really sent me off. Unless one really can add something
> useful, helpful, enlightening or usefully provocative, there really isn't
> much point posting to a thread, I think.

I start to understand you, it's probably similar to a feeling I got the other day when I found my name in Nial's Oracle weblog. I had to read and reread what he wrote and the accompaning material until I could convince myself that he wasn't criticising me and cleared the comment I was writing.

And to let this whole thread go completely astray, Nial quoted a metalink note but left out a newline between two sentences which made it really hard to see what he was getting at. By just ommitting a newline, he gave the whole paragraph a slightly different meaning, not on purpose though.

It's not really interesting, but for those interested you can find what I mean here (August 31, 2004):

http://www.niall.litchfield.dial.pipex.com/index.html

>
> Look at the current 'can you make a table read-only' thread for an example
> of useful, constructive, provocative "thread drift". Joel's effort isn't
> anywhere in that league.
>
>
> But whatever.
>
> In this Usenet world we live in, we have to put up with much worse than Joel

Yepp. There are still some eggs to fry ;-)

> from time to time. And I am over the entire discussion, especially since
> the OP hasn't come back to enlighten the discussion any further (and who
> can blame him -though that's another reason why I felt Joel's post was
> completely unwarranted and unnecessary).

I'm sometimes wondering why OP's don't come back, but probably this 'grab and run' behaviour is what you're getting today. Serves'em right! :-)

>
> Your thoughtful comments have indeed been worthwhile. I hope my response
> shows some measure of similar thought: that I'm not just over-sensitive or
> paranoid.

I feared since I've been commenting you a lot recently, you might start to think I've go on you. I don't, so your answer is much appreciated.

Cheers

Holger Received on Wed Oct 20 2004 - 05:39:16 CDT