All 'areacello' data failed tests.

[jbusecke]

I just noticed when looking for the variable areacello that all the legacy stores are in out non-qc catalog.

I suspect that this is an error in the tests (likely some assumption about time being in the dimensions).

• This will require some way to efficiently rerun test over all datasets (except retracted ones) after adjustments have been made to the testing.

[jbusecke]

Yeah this seems like it would fail if there is no time present in the dataset.

[briochemc]

+1 for areacello!

It should be OK to only test "failed" data if a PR fixing this issue only removes tests (skips the time test if there is no time dimension), right? (I'm just asking, I may be confused as I don't really understand much of how this all works 😅)