Dataset refresh parallelism

[FredrikJ]

Hi!

Our datasource is a Orcle database and I am trying to figure out why the first partition never is run with parallelism when refreshing a dataset.

If I run 7 partitions in a table the parallelism looks like this:

The first partition runs on its own then the other 6 in parallelism

If I run 5 partitions in a table the parallelism looks like this:

The first partition runs on its own then the other 4 in parallelism.

Does any one know why not all 5 partitions runs in parallelism in my last run?

[GilbertQ · Accepted solution]

Hi @FredrikJ 

This is by design because the first query that runs, does the checks to ensure that the columns are valid and that there are no refreshing errors. Once the first query is successful it can then refresh in Parallel knowing that the subsequent queries will not fail.