tripal/t4d8

Handling multiple Chado schema

Closed this issue · 1 comments

This issue is there to address different points than the one in #52.

Here, we should discuss technical issues related to handling several Chado instances. Previous Tripal version were working on a single instance. While the Multi-Chado module allowed to trick Tripal and change the "active" Chado instance, it was not able to work simultaneously with several instances.

Tripal 4 brings the opportunity to work with several data sources (either Chado or not) at the same time. But we have to define the behavior of our system. Questions are:
Do we want to have a default Chado schema or each time something is using Chado, it must specify which schema?
Should we want to manage multi-schema SQL queries? (what are the use cases? what could we imagine?)
Should we handle Chado schemas in different databases? ...which means 2 Chado instances could share the same schema name but reside in 2 different databases. --> then a Chado instance should be specified by both its schema name and database name.
How would this work with Drupal views?
How would this work with Drupal entities? A same entity type could be found in several Chado instances.

It seems complex but there are technical solutions to those problems. We could also use a smart approach: implement current Tripal version for a single "active" chado instance while adding all the necessary stuff to work with several instances and implement their use later and/or enable them through a setting (like "[ ] enable multiple Chado instances (experimental)").

This has been solved in tripal_biodb branch (#134) on the Chado management admin interface. A default schema can be selected there.
Closing issue.