Should pool names be unique?

[wasade]

Pool names are not enforced to be unique. In addition, the same pool can be re-pooled. A caveat of the combination of these is that the same pool name can be added to the same pool name multiple times.

[charles-cowart]

High priority to at least resolve the question. My own thinking would be yes; this seems similar to plate names effectively being unique, even though they aren't constrained as such in the schema.

[charles-cowart]

Asking @ghsmu414 to weigh in.

[ghsmu414]

It seems like there are a few issues with this step. You should not be able to pool from the same pool for the same sequencing run. I am fine with a unique ID for each pool because when we have to run a pool again, we can rename it. 16S_Knight_Pool1 to 16S_Knight_Pool1_ReRun1

But I am not sure I am answering the question fully.

[wasade]

Thanks, @ghsmu414!

[charles-cowart]

The natural thing to do would be to add a uniqueness constraint on the pool's name in the database. Unfortunately, it appears that a pool's name is not stored in labman.pool or even labman.composition, but is stored in 'external_id' in labman.tube. Not knowing what else external_id in labman.tube might be used for, adding a constraint to that field is a non-starter. I also think that enforcing some kind of uniqueness at the GUI level is bad design.

My suggestion is to have an actual column in labman.pool named appropriately, and refactor the queries.