Hi,

There is no way that I'm aware of to just trigger one task in a pipeline. You can repair a run, but this will just trigger everything in the job from the point it failed out. If I've understood you correctly, the better approach may be to separate your table creations into individual jobs, then run an overall job to execute them for your day-to-day schedule. You will then be able to execute the individual components as well, as and when you need to.

I hope that helps.

Thanks,

Emma