Reputation: 1
I'm currently facing to a datafactory problem and can't managed to find a solution: I've mad a pipeline with a copy activity which send data from a datawarehouse to an oracle database.
The pipeline run fine, but take a lot of time (4h-5h). I think it's caused by the amount of data (Approximately 12 millions rows)
Do you have any idea for making this pipelin run faster?
Thanks for your help,
Guif
Upvotes: 0
Views: 102
Reputation: 102
Try bigger DIU (old name: DMU) and increase the parallelism in the copy activity.
You may also try stage the data into a blob first then followed by a copy to Oracle. I know SQL to SQL would have better performance with staging.
Be sure to use an IR region that is close to your data sources.
Upvotes: 1