A number of virtual machines in a vSphere 5.5 DRS cluster are connected to the same virtual switch port group. After several of these VMs are migrated to another host, an administrator notices that th

Network I/O Control manages bandwidth allocation and traffic shaping but does not limit the number of ports available on a virtual switch, so it would not produce an 'Out of resources' port exhaustion error.

DRS automation level and migration threshold control when migrations are triggered, not port capacity on the destination host's virtual switch - changing this setting would not cause port exhaustion.

Insufficient uplinks would reduce available bandwidth and potentially cause packet loss or failover events, but would not produce the 'Out of resources' port count exhaustion error seen in the vmkernel log.

In vSphere 5.5, the default virtual switch port allocation is elastic, meaning ports are created on demand. If an administrator changed the allocation to a fixed number that is lower than the total number of virtual machine NICs that need to connect post-migration, the switch exhausts its available ports and the vmkernel logs 'Out of resources', preventing the vNIC from connecting to the port group.