I created this initial task as a starting place but really have no idea what the scope of work is for this so please edit / claim ownership boldly :)

Let me know if I can help on anything, one note would be that I hope these changes would make computation easier. For example if you're building the list of most used templates, you can do all the work with numeric ids first and at the end just translate the final result from id to strings. That would reduce memory usage and computation load drastically.

Thank you @Isaac for creating this task!
I have reviewed the planned change and it will indeed impact our sqooping/usage of the various links tables.
I've added myself as a subscriber of the migration task to follow implementation - Would you have any precision in term of implementation timeline @Ladsgroup?
I'll start planning the needed changes on the Data-engineering side early next week.

In T304979#7883363, @JAllemandou wrote:

I've added myself as a subscriber of the migration task to follow implementation - Would you have any precision in term of implementation timeline @Ladsgroup?

It's a bit hard because it depends on the wiki. For some wikis it's not started yet and for some wikis it's already finished (it's going to be write both for a while though). For some wikis it took an hour to finish for some it'll take months.

You can easily check if it's finished or not by a query like:

select * from templatelinks where tl_target_id is NULL limit 1;

And if it returns anything, it means it's not done yet.

Thank you @Ladsgroup for the update.
We need to prioritize the change in DE soon :)

Ping @EChetty - This task needs to be priositized by the team - the SQL change is already happening and will impact some of our jobs.

@EChetty moving this to DE planning backlog because

This task needs to be prioritized by the team - the SQL change is already happening and will impact some of our jobs.

Thanks @JAllemandou