Open louismartin opened 2 years ago
Yep, my take was that the marginal value of garbage collection compared to the complexity it adds to the code wasn't worth it. Keep in mind there is no real DB just files over NFS. If you have an easy and reasonably reliable solution I would be happy to take it.
I had a grid
grid_a
with too many experiments running so I refactored some of its experiments in a new grid filegrid_b
. While running the new griddora grid grid_b
worked as expected and found the already running experiment, when I randora grid grid_a
again it cancelled all the experiments that were now ingrid_b
.It would be nice to have a way to track this scenario and only garbage collect experiments that are not linked to a grid. And also it would be nice to ask the user for confirmation when cancelling experiments.