You can customize the default policy and base policies included with vRealize Operations Cloud for your own environment. You can then apply your custom policy to an individual object or groups of objects, such as the objects in a cluster, or virtual machines and hosts, or to a group that you create to include unique objects and specific criteria.
You must be familiar with the policies so that you can understand the data that appears in the user interface, because policies drive the results that appear in the vRealize Operations Cloud dashboards, views, and reports.
To determine how to customize operational policies and apply them to your environment, you must plan ahead. For example:
- Must you track CPU allocation? If you overallocate CPU, what percentage must you apply to your production and test objects?
- Will you overallocate memory or storage? If you use High Availability, what buffers must you use?
- How do you classify your logically defined workloads, such as production clusters, test or development clusters, and clusters used for batch workloads? Or, do you include all clusters in a single workload?
- How do you capture peak use times or spikes in system activity? In some cases, you might need to reduce alerts so that they are meaningful when you apply policies.
When you have privileges applied to your user account through the roles assigned, you can create and modify policies, and apply them to objects. For example:
- Create a policy from an existing base policy, inherit the base policy settings, then override specific settings to analyze and monitor your objects.
- Use policies to analyze and monitor vCenter Server objects and non-vCenter Server objects.
- Set custom thresholds for capacity settings on all object types to have vRealize Operations Cloud report on workload, and so on.
- Enable specific attributes for collection, including metrics, properties, and super metrics.
- Enable or disable alert definitions and symptom definitions in your custom policy settings.
- Apply the custom policy to an individual object or groups of objects.
When you use an existing policy to create a custom policy, you override the policy settings to meet your own needs. You set the allocation and demand, the overcommit ratios for CPU and memory, and the thresholds for capacity risk and buffers. To allocate and configure what your environment is actually using, you use the allocation model and the demand model together. Depending on the type of environment you monitor, such as a production environment versus a test or development environment, whether you over allocate at all and by how much depends on the workloads and environment to which the policy applies. You might be more conservative with the level of allocation in your test environment and less conservative in your production environment.
When you establish the priority for your policies, vRealize Operations Cloud applies the configured settings in the policies according to the policy rank order to analyze and report on your objects. When you assign an object to be a member of multiple object groups, and you assign a different policy to each object group, vRealize Operations Cloud associates the highest ranking policy with that object.
Your policies are unique to your environment. Because policies direct vRealize Operations Cloud to monitor the objects in your environment, they are read-only and do not alter the state of your objects. For this reason, you can override the policy settings to fine-tune them until vRealize Operations Cloud displays the results that are meaningful and that affect for your environment. For example, you can adjust the capacity buffer settings in your policy, and then view the data that appears in the dashboards to see the effect of the policy settings.