[pas]

> But how do we test that our migrations behave correctly in the face of long-running transactions? I.e. what's the failing test case for that bug?

Isn't it enough to simply make sure the migration transaction successfully finished? Even if there is a long running transaction, if the migration finished, that long tx will get aborted and rolled back.

Or if the migration stalls because the long running tx, then you'll presumably get a timeout error.

Is there something I'm missing?