DSXchange

Hi All,

I was trying to testify Partition Table option in oracle Enterprise stage for performace tuning, I wrote a simple select query and insert to a dataset
the issue is that the number of extracted rows got doubled after I gave the source table as a partition table, I am sure there are several indexes sitting in the souce table.
I had some study with previous posts regarding this topic, still didn't get the point.
Please give any advice, and what is the senario we can use this option(one and only one index in source table?)
thank in advance and your time is highly appreciated!

Yan

Is your source table in fact partitioned? Note that index <> partition, you seem to be confusing the two. Check with your DBA if you are unsure.

Thanks so much Craig,
Sorry for my green , do you mean only table itsself is partitioned, we can use this option, and index has nothing to do with this?

Well... there will be specific indexes that support the partitioning, but yes - only if the table itself is partitioned.

Hi,

Just to confirm,

Is there any special requirement for oracle database to use the Partition Table option in oracle Enterprise stage?

As per craig,

chulett wrote:Well... there will be specific indexes that support the partitioning, but yes - only if the table itself is partitioned. ...

there are special requirements for oracle database, at least, "only if the table itself is partitioned", however, replies in this thread
viewtopic.php?t=91074 seem otherwise.

Anyone can confirm this?

thanks