Re: Locally managed tablespaces and migrated rows

By the way - a note for Howard particularly. Some time back I posted a note that you
still get dictionary activity if you do something like offline, and/or make readonly, an LMT then drop a segment from it.

I think I said that the dropped extents are recorded in uet$ until the tablespace is brought online and read write. This should have said SEG$ - each dropped extent becomes its own segment. However, this doesn't always happen, and I can't figure out why it happens sometimes - I think it must be a very specific timing consideration that is, perhaps, unlikely to happen under any sort of normal usage.

--
Jonathan Lewis
Yet another Oracle-related web site:  http://www.jlcomp.demon.co.uk

Practical Oracle 8i:  Building Efficient Databases
Publishers:  Addison-Wesley

Reviews at: http://www.jlcomp.demon.co.uk/book_rev.html



Jonathan Lewis wrote in message
<989397070.12095.0.nnrp-10.9e984b29_at_news.demon.co.uk>...


>

>I think your argument can be summed up

>very briefly -

>

>If you know exactly what you are doing

>there is very little advantage to LMTs

>over DMTs.  But with a little care and

>ingenuity it's quite easy to turn a safety

>feature into a death trap, particularly if

>you call it a shrink-wrapped, black-box,

>3rd-party application.

>

>

>--

>Jonathan Lewis

>Yet another Oracle-related web site:  http://www.jlcomp.demon.co.uk

>

>Practical Oracle 8i:  Building Efficient Databases

>Publishers:  Addison-Wesley

>

>Reviews at: http://www.jlcomp.demon.co.uk/book_rev.html

>

>

>

>Nuno Souto wrote in message <3af7b8a4.4414731_at_news-server>...

>

>>

>>Anyways, that's the core of the problem I had.  Education being

>>perceived as promoting the thing as a "blanket".  It's not always that

>>black and white.  Most of the time it's bloody dark grey...

>>

>

>

>