On ESX hosts, DHCP failures might occur in the service console during heavy loads
Under particularly high stress situations, the service console may not have sufficient resources to renew a DHCP lease. For example, the following can occur:
The service console is not scheduled due to excessive processor loads.
A high network packet rate creates enough pressure to prevent DHCP responses from reaching the service console.
The DHCP client will only retry lease renewals a limited number of times. If it cannot renew its DHCP lease, the connection to the service console is lost.
The best practice is to assign a static IP address to the service console when feasible and to refrain from running applications with dramatically high workloads, which can overload the service console.