You are getting this error because when you activate the persistence, the baseline doesn't adjust by itself anymore.

So the data you stored into your caches are lost when the only node in the baseline shutdown.

You need to enable the baselineAutoAdjustEnabled so that the node leaving or joining the cluster will enter/leave the baseline and share all the data of your cache.

You should also configure backup to your cache and set the mode to REPLICATED to be sure you won't lost data.

Hope it help