Is this esx a part of a drs cluster? What is the cluster DRS setting set to? Fully automated? What about aggressive level?
There can be a couple of reasons for this
1. DRS set to manual or disabled. Load distribution would need to be done manually. Move a couple of memory aggressive VM's from this host to any other and check
2. Restart the management agents on the host (services.sh restart) and check
If DRS is already enabled, try "disconnecting" the host from VC (only disconnect not remove) and connect it back and check