Continuous optimization provides ongoing, autonomous management of vRealize Automation workloads by vRealize Operations Manager.
With continuous optimization, you take advantage of workload rebalancing and relocation, and use vRealize Automation with vRealize Operations Manager beyond initial workload placement. As virtualization resources move or come under heavier or lighter load, vRealize Automation provisioned workloads can move as needed.
- Continuous optimization automatically creates a new datacenter in vRealize Operations Manager.
There is one new datacenter for each vRealize Automation vCenter endpoint.
- The newly created datacenter contains every vRealize Automation managed cluster associated with the endpoint.
Note: Do not manually create a mixed datacenter of vRealize Automation and non- vRealize Automation clusters.
- You can run continuous optimization only from the newly created vRealize Automation based datacenter.
- Optimization does not support different reservation requirements among clusters in the vCenter, which might occur when you have different business groups.
Optimization is at the vRealize Automation based datacenter level, and different reservation requirements across clusters can prevent success. When that happens, an error appears, stating that some destination clusters or storage did not meet requirements, which prevented some optimization moves.
- Optimization never creates a new vRealize Automation or vRealize Operations Manager policy violation.
- If you have existing policy violations, optimization can fix vRealize Operations Manager Operational Intent issues.
- If you have existing policy violations, optimization cannot fix vRealize Operations Manager Business Intent issues.
For example, if you manually moved a virtual machine to a cluster that was not a part of its reservation policy, vRealize Operations Manager does not detect a violation nor try to resolve it. To fix Business Intent issues, you must use vRealize Automation to move the workload.
- This release obeys Operational Intent at the datacenter level. All member vRealize Automation clusters are optimized to the same settings.
To set a different Operational Intent for clusters, you must configure them in separate vRealize Automation datacenters, associated with separate vCenter endpoints. Having different test and production clusters might be one example situation.
- vRealize Operations Manager queries vRealize Automation for allowable placement based on vRealize Automation policies and reservations.
- vRealize Operations Manager placement tags cannot be applied to vRealize Automation provisioned workloads.
In addition, scheduled optimization involving multiple machines is supported. Regularly scheduled optimizations are not all-or-nothing processes. If conditions interrupt machine movement, successfully relocated machines stay relocated, and the next vRealize Operations Manager cycle attempts to relocate the remainder as is usual for vRealize Operations Manager. Such a partially completed optimization causes no negative effect in vRealize Automation.