Oracle FAQ Your Portal to the Oracle Knowledge Grid

Home -> Community -> Usenet -> c.d.o.server -> Re: Direct path write

Re: Direct path write

From: Howard J. Rogers <>
Date: Wed, 18 Aug 2004 16:07:18 +1000
Message-ID: <4122f235$0$11790$>

Jaap W. van Dijk wrote:

> On Tue, 17 Aug 2004 09:33:53 +1000, "Howard J. Rogers"
> <> wrote:

>>Jaap W. van Dijk wrote:
>>> On Mon, 16 Aug 2004 21:59:56 +1000, "Howard J. Rogers"
>>> <> wrote:
>>> ...
>>>>Try not rebuilding indexes at all, since they almost certainly don't
>>>>need it.
>>> Maybe not in a transactional application.
>>> For bulk loads they can have been set to UNUSABLE.
>>> Jaap.
>>Come on... I said "almost". Inevitably, there are circumstances where they
>>must be rebuilt (alter table move, anyone?).
>>Bulk loads are just one amongst many.
>>The point was not to iterate all possibilities, but to get the OP to
>>validate the real *need* to rebuild before going ahead and doing it.
> The picture just struck me as funny: this guy having a bunch of
> unusable indexes and somebody advising him not to rebuild them at all.
> Clearly that was not what you had in mind.

Unless you'd care to postulate that I am a complete moron and know zilch about the Oracle RDBMS, it is highly unlikely that anything even remotely along those lines would be what I had in mind.

Of course, I can't stop people reading what I write with a bloody-minded literalness that makes the words mean nonsense.

But even if you insist on doing that, try reading ALL the words before parsing them. The 6-letter one "almost" has a significance in the sentence you apparently missed.  

> Maybe we can agree on 'Try not rebuilding valid indexes at all, since
> they almost certainly don't need it'.
> Jaap.

No we can't agree on that, because it would imply the readers here are morons too. Leave it at my original formulation, and trust to the innate intelligence and good sense of most DBAs to spot an exception to the general rule when it arises.

HJR Received on Wed Aug 18 2004 - 01:07:18 CDT

Original text of this message