Data loading is slow not using direct path

[desmondchew]

We setup a new task, got the following INFO.

How can we use direct path which is faster?

Both source and target endpoint is Oracle 11.2.0.4.

Table 'ABC' has an index, SQL will be used for the load process instead of the Direct Path

We only have one table in the task. We configured parallel load with partitions and set a filter column. The filtered column DATE has an index in the source database.

Desmond

[desmondchew]

Uploading screenshot showing the no direct path.

[john_wang]

Hello Desmond, @desmondchew 

This appears to be a configuration issue. The option "Use direct path full load" cannot be used together with Replicate parallel load. This is because the Oracle direct path load utility locks the target table during the load operation. As a result, when one loading thread locks the table, it prevents other threads from loading data simultaneously.

Therefore, Direct Path Load and Parallel Load cannot be used at the same time, you can only enable one of these options.

Hope this helps.

John.

[desmondchew]

Hi John,

What would you recommend us to use direct path or parallel load? Which is faster?

Thank you.
Desmond

[john_wang]

Hello Desmond, @desmondchew 

It is difficult to determine at this stage, as the performance may be affected by several factors, including network bandwidth, the level of parallelism, and the data volume etc.

I would suggest starting with parallel load and performing a performance comparison test to evaluate the results.

Thanks,

John.