0 votes
One of my dataset is partitioned along two dimensions (source and date), but I'd like to retrieve a non-partitioned dataset from it. When I try to run a simple "Sync" recipe using "All available" partitions, the build fails systematically, probably because I don't have a source partition for each existing date. Is it possible to find a way around this?

1 Answer

+1 vote
Best answer

A fairly ugly but efficient way of doing that is to define another dataset, non partitioned, pointing to the same location (table or file path). However, there won't be a link in the Flow for that.
selected by
Yep, that's ulgy indeed, but it works. Thanks!
1,296 questions
1,325 answers
11,862 users

©Dataiku 2012-2018 - Privacy Policy