Re: Migrate from DMT to LMT when fet$ and uet$ counts > 700,000

Charles Davis wrote:

>
> Thank you for the reply.
>
> Can you please elaborate on what you mean by a steaming mess, in technical
> terms.

No. I don't do technical terms. What I meant was already spelled out in my last post.

> I believe I can not simply reorganise this database in LMTs to correct the
> problem; I believe I need an interim step to alleviate the problem.

Stop believing things. Oracle works on science and facts, not faith and belief.  

> I believe that if I were to reorg into proper LMTs now, then the final
> drop
> of the object in the DMT will only increase the coalescing problem.
> Oracle will convert the dropped object to a temporary object and them SMON
> will
> have to deal with it as a DMT object and coalese it. That seems to me to
> only exacerbate the problem.

And since when does SMON "have to deal with" coalescing? SMON only automatically coalesces when PCTINCREASE <> 0. So take charge of your database, instead of relying on automatic convenience, and it won't be a problem.  

> Please tell me, in as detailed technical terms as you can, what mess will
> be created by Migrate-to-local prior to me properly reorging the 1.8TB
> database into proper LMTs.

I already did. The procedure doesn't even attempt to sort out extent sizes or tablespace fragmentation. If you have a steaming mess of fragmentation before, it will still be steaming afterwards. It fixes up nothing.

If you're going to use MOVE to fix the problem up anyway, then do so. You can deal with SMON coalescing things as part of that process simply by telling it not to.

> This is generally what I originally asked.

And it's specifically what I originally answered.

HJR Received on Sun Oct 24 2004 - 14:47:05 CDT