Remember when you thought vRealize Operations 7.0 was a giant release packed with goodness? That was just the warm up. Welcome to vRealize Operations 7.5 which contains so many new and improved features I will have to break this technical overview blog into four chunks, so you don’t break your mouse’s scroll wheel in one go.
Why four? Well, with the release of 7.5 we are matching the four tenets of Self-Driving Operations with the four pillars of use cases found on the Quick Start screen.
Each of these areas is gaining something new and exciting with 7.5, so I will break down this blog into a series that covers each of them, plus a bonus post on “The Best of the Rest” to round out things that help with overall usability which are still important things you should know.
Let’s get started with the first pillar, Workload Optimization. If you need a review of what Workload Optimization is and how you can benefit from it, check out these demos!
vSAN Aware Placement
Virtual machines running on vSAN datastores can now benefit from Workload Placement in vRealize Operations. This means Workload Placement can now migrate virtual machines between vSAN clusters to enforce Operational Intent (balance, moderate or consolidate virtual machines) and Business Intent (placement on hosts or clusters based on vSphere tagging). Specifically, the Workload Optimization is respectful of three aspects of vSAN that influence placement decisions:
- Storage Policy
- Slack Space
- Resync
Each of these is considered during a workload optimization event to insure to any moves will enhance workload performance and not create any bottlenecks.
As you know, vSAN requires each virtual machine to be assigned a storage policy. This provides a set of basic storage capabilities related to data availability and protection as well as capacity reservations. Workload Placement relies on vCenter to validate each virtual machine’s storage policy against all available target vSAN datastores. Any vSAN datastores which can support the VMs storage policy are considered candidates for migration destination.
So, what is Slack Space? Since vSAN is a distributed storage solution, it needs free space to perform these actions that are transparent to the user. Any of the changes listed will use this slack space temporarily to adjust to the new conditions. However, it cannot be deemed as capacity always used for temporary actions, since the resynchronizations may have occurred the result of the user prescribing a higher level of resilience. As such, vRealize Operations now understands how much slack space is needed by a vSAN datastore so it can be subtracted from the overall space available.
Finally, a resync can occur any time vSAN determines the need to do so. It protects the availability of data in accordance with virtual machine storage policies. This can create additional network IO on the cluster. Given that this important operation should not be interrupted or impacted by migrating virtual machines, Workload Optimization will avoid migration of virtual machines to vSAN clusters running resync.
A couple of things to keep in mind when using Workload Optimization in a vSAN environment:
- In order to get the full benefits you will need to enable the vSAN mgmt pack. That’s easy to do because it’s native inside of vROps.
- Workload Placement will not migrate a virtual machine from vSAN to a datastore cluster or vice-versa (e.g. it will only move a vSAN VM to another vSAN cluster).
Host Group Visualization of Usage
In 7.5 we have enhanced the Workload Optimization cluster details to show host groups created through the Business Intent settings. For example, if you use Business Intent to enforce placement of virtual machines on specific hosts for license cost reduction, Workload Optimization automatically creates host groups in DRS (as well as virtual machine groups).
Now, you can dig deeper into cluster utilization to see how each of the host groups is using CPU and memory.
For example, in the screenshot above there are two clusters in this datacenter. The datacenter itself is set to an Operational Intent of “Balance” and a Business Intent for Host Based License Enforcement. The clusters themselves are balanced well, however that does not mean everything is running smoothly. Notice that the Linux host group on cluster 02 is showing very high CPU workload, while the Linux host group in cluster 01 has much lower CPU workload. The datacenter needs to be optimized to alleviate this high workload on the Host Group. Running Workload Optimization will migrate Linux virtual machines from cluster 02 to cluster 01, ensuring they stay within the appropriate Host Groups and honoring license enforcement.
If a cluster is constrained and cannot meet Operational Intent due to restrictions placed by Business Intent (for example, both Linux host groups above are CPU constrained), then you can see where the bottleneck is within host groups and decide which of them needs more resources (then you can run a What-If scenario to plan for more hosts).
Simplified Violation Visualizations
Along with host group visibility, you can now get more specific detail on Business Intent tag violations. Previously, in Workload Optimization, you would only see the tag involved in the violation at the datacenter level. Helpful, but it would be better to know which objects are triggering the violation. Now, you can easily find out which virtual machines aren’t where they need to be according to your Business Intent.
Returning to our license enforcement use-case for Business Intent, the screenshot above shows there is a virtual machine tagged with “Oracle”, yet it is running on the host group for Linux. This gives you quick and easy understanding of the details that caused the tag violation instead of waiting until you run the optimization to view the moves. And in the event that the virtual machine cannot be moved due to resource constraints (i.e. the host group is at 100% utilization for CPU or memory) you will at least know the impact to your Business Intent to help justify any additional purchases or reclaims that need to be made.
Try It For Yourself
A good way to learn more is to download a trial of vRealize Operations 7.5 and try it out! You can also find more demos and videos on vrealize.vmware.com. Be sure to check out the next blog in this series where I will cover what’s new in 7.5 in Optimize Capacity, including the return of the allocation model, discovery and reporting of orphaned disks, new What-If scenarios and more.
