Oracle and in-database processing

apivk

Hello,

Is anyone aware of an alternative on executing a visual query as in-database processing on Oracle?

Found this In-Database Processing extension (marketplace.rapidminer.com/UpdateServer/faces/product_details.xhtml?productId=rmx_in_database_processing), but seems it only supports Google BigQuery (via OAuth 2), PostgreSQL and MySQL.

Ideas welcome.

BalazsBarany

Hi @apivk,

support for additional databases can be added by RapidMiner developers if a customer is asking for it. This hasn't been the case yet for Oracle, it seems.

That said, you can just try it. SQL is reasonably well defined, so most things (SELECT from tables, JOIN, filter, aggregate, union) should work nevertheless. Generate Attributes won't help you with the available functions but you can still enter SQL expressions.

Regards,
Balázs

tkenez

Hey @apivk ,

Some more thoughts on this. As Balazs said above, we don't yet support Oracle in In-Database Processing Extension. It is on the roadmap, but I cannot say a firm date for it just yet.

In the meantime, if you have the skills to meddle with Java code, there is a sample project on GitHub on how to extend the extension to support new databases, and Oracle is used in the example. Check it out here: https://github.com/rapidminer-labs/rapidminer-extension-indb-oracle-sample

The reason why this hasn't been simply added to the extension just yet is that there are some kinks to work out so we can safely say it can support production workloads.

BalazsBarany

Hi @apivk,

support for additional databases can be added by RapidMiner developers if a customer is asking for it. This hasn't been the case yet for Oracle, it seems.

That said, you can just try it. SQL is reasonably well defined, so most things (SELECT from tables, JOIN, filter, aggregate, union) should work nevertheless. Generate Attributes won't help you with the available functions but you can still enter SQL expressions.

Regards,
Balázs

tkenez

Hey @apivk ,

Some more thoughts on this. As Balazs said above, we don't yet support Oracle in In-Database Processing Extension. It is on the roadmap, but I cannot say a firm date for it just yet.

In the meantime, if you have the skills to meddle with Java code, there is a sample project on GitHub on how to extend the extension to support new databases, and Oracle is used in the example. Check it out here: https://github.com/rapidminer-labs/rapidminer-extension-indb-oracle-sample

The reason why this hasn't been simply added to the extension just yet is that there are some kinks to work out so we can safely say it can support production workloads.