Sign up to take part
Registered users can ask their own questions, contribute to discussions, and be part of the Community!
I imported a project on my Dataiku instance but I have some problems with two of the six input datasets, for which I obtain the error "the root path for this database does not exist". Could you please help me? What's the problem? I also contacted the creator of this project but I didn't solve the situation.
Thank you so much
where are your input datasets coming from ?
Looks like they are stored in a database. If you are moving projects to a different DSS, you need to make sure you have the same connection on your target instance (it can have a different name, though, you will remap it upon project import).
Otherwise, DSS won't be able to retrieve data. Another option is to include the input datasets when exporting the project, however this might be an issue if the datasets are very big.
Also consider that if you are using one or more connections in your project (thus you are reading or storing data on them), DSS will need to find a matching connection in the target instance as well (it can have a different name, but has to be the same kind), or you won't be able to import the project.
Hope that helps.
Architect @ Dataiku
datasets are not connected to a database, they are csv files, but I think that those without an existant root path were created previously in another project as a result of a previous workflow. I don't know if this problem could be related to this fact. I don't understand why the connection to data source is ok. I send you a screen of my situation
Thank you for you help!
you mention that the connection is ok and those datasets might have been created by another project.
When you create the export bundle for this project, can you select to embed those datasets as well (if they are not too big)?
Otherwise I recommend, on the original DSS, to run a sync to move the datasets to a connection that you know will be available in the target machine as well.
On the target machine, you can drop the original datasets, only keeping the sync copy
If none of this works, please open a support ticket.
Architect @ Dataiku