Configuring overcommit rates

The capacity in the IBM Cloud Manager - Self Service user interface indicates the total and used (or allocated) resources, including processors, memory, and storage.

All x86 clouds support resource overcommit by the hypervisors. Therefore, the total amount of virtual resources that can be allocated are larger than the total physical resources available. So, the IBM Cloud Manager - Self Service user interface supports the ability to set an overcommit rate to limit the resource use.

Note: Overcommit rate is not supported for PowerVC virtualization environments.

The overcommit rate is represented in the following fields:

totalCpu - Represents the physical processor cores that are multiplied by the processor overcommit ratio
totalMem - Represents the physical memory size that is multiplied by the memory overcommit ratio

Note: Overcommitted storage is not allowed.

The capacity API tells administrators the amount of physical resources, the amount of resources (after overcommitted), and the resources that are allocated. The user interface shows the overcommit rate only. It excludes the amount of resources after they are overcommitted.

To configure the overcommit rate, open the cloud.properties file and set the following properties:

# The cpu overcommit rate in OpenStack cloud
com.ibm.cfs.openstack.overcommit.cpu=16
# The memory overcommit rate in OpenStack cloud
com.ibm.cfs.openstack.overcommit.memory=1.5

com.ibm.cfs.openstack.overcommit.memory is invalid for hyper-v hypervisor.

The IBM Cloud Manager - Self Service backend checks whether there are sufficient available resources when deploying. If the available resource is less than the requested resource, the deployment process stops. This feature can be enabled or disabled by using the following property in the deployment.properties file:

#Enable/Disable the capacitiy check against the selected target while deploying a workload.
com.ibm.cfs.resource.check.enable=true