where can the we configure the system health trsholds for stale collections

NNMi 23.4

Linux-Standalone

Customer gets all the time system help messages: Stale collections (8) has status Major because there were between 5 and 10 stale collections in the last 5 minutes.

Some of the nodes are not responding at all, and the customer has 4000 nodes, that there are stale collections is normal, but where can we change the threshold's for the number and time span of this messages in NNMI?