Exactly, thanks, Alberto! But in general, is it best practice to restart a cluster every week to prevent this issue? Or does this problem happen because the code is not well-written?