Re: newbie plsql

The free extents size after i drop the index , are not the problem . First i fill the tablespace where the index is , then i drop the index and recreate it .
If i drop the index in a block plsql and after , in the same block for example , select free extents in that tablespace ,they aren´t there . But if i execute the create index instruction in sqlplus as it is in the block plsql , i create it exactly with the same extents .

I´m trying to defrag data that is insight blocks not extents of a segment ( index stagnation ) .

Roger Snowden wrote:

> The 'commit', or lack thereof, is not the problem. Any DDL, such as
> 'create' or 'drop' will invoke an implicit commit for the session.
> What is
> probably happening is that the original index grew incrementally with
> multiple extents. You might have set your storage parameter to have an
>
> INITIAL extent to hold the entire thing and you don't have enough
> contiguous blocks to do so. You might cut back on the storage
> parameter and
> allow the index to extend as it gets created. Doesn't help the defrag
> though, does it? So... you must export all objects in the tablespace
> and
> defrag the tablespace. Export objects, drop tablespace, import
> objects.
>
> Roger
Received on Fri Oct 10 1997 - 00:00:00 CDT