By default, system services are started on one management host until all the CPU slots on the host are used, then another host is selected and services are started on that one until all the CPU slots on that host are used, and so on. This can cause some management hosts to be overloaded while other hosts remain idle.
You can, however, configure the system so that system services are evenly distributed on management hosts. You do this by configuring the Balanced slot allocation policy on the ManagementHosts resource group.
When the Balanced slot allocation policy is enabled, hosts are selected according to the number of free CPU slots on the host. Since the number of free CPU slots on a host decreases with each allocation, the same host will not be reselected unless it has the highest number of free CPU slots.
When you change the slot allocation policy, the new policy is only applicable to allocations requests made after the policy was changed.
In some cases, you may want to apply the policy immediately so that system services are redistributed equally among management hosts. You can do this by stopping and restarting the system services. Note that restarting services will affect cluster operations. Refer to the table below for details.
The following is a list of out-of-box services and the effect on your cluster during service restart: