Hi Rob – apologies for the (very) belated reply. Bottom line – agreed – I think it is a balancing act. You might want, for example with a large data load, reduced checkpoints so that the data load checkpoints don’t compete. But then the checkpoints that do happen could be far more pronounced (and longer in duration). So it depends on the goals. Ideally we can keep the defaults, but there may be scenarios that justify the custom checkpoint frequency.