Formally Known as vCenter Operations (vCOps), vRealize Operations Manager (vROps) really has become the center of the vSphere universe. vROps is an appliance that sits in your environment collecting system metrics from vCenter, virtual servers and ESXi. It acts as the single pane of glass to the virtual environment, allowing the administrator to track and mitigate resource contention, along with performance and capacity constraints. vROps will also “learn” about the environment, and given a couple of months, the data collection can be used to perform future calculations to determine things like when more capacity is required based on growth and resource consumption, and that’s pretty cool. Data collection is not just limited to VMware products however, you can also install additional management PAKs from VMware’s Solutions Exchange, and there are always more being added. Which brings us to the next topic: Sizing your vROps deployment.
Unlike vCOps which consisted of two virtual machines within a vApp, vROps is a single virtual appliance, that can be expanded and clustered for additional compute resources. A single appliance or node can be deployed and looks like the following in figure 1:
Figure 1 – A single vROps node
Now the cool thing about vROps is that it has the built in functionality of clustering two appliances together as a Master and Master Replica, giving you resiliency in case of a failure. You can also add additional nodes to the cluster known as Data nodes, that will allow you to collect and process even more metrics. It should go without saying, but as you add more Management PAKs from VMware’s solution exchange, keep in mind that you may need to add additional data nodes. It’s also important to note that the master and master replica servers can also be referred to as data nodes, and that is important because since vRealize Operations Manager 6.x, you can have a total of 16 data nodes in a cluster. That means you can have an HA pair, and 14 additional data nodes. You can deploy the appliance in several sizing configurations depending on the size of your environment and those are: Extra Small, Small, Medium and Large.
- Extra Small = 2 vCPUs and 8GBs of memory
- Small = 4 vCPUs and 16GBs of memory
- Large= 16 vCPUs and 48GBs of memory
You can also deploy additional appliances known as a Remote Collector, and since vRealize Operations Manager 6.1.x, you can have a total of 50 remote collectors, allowing the collection from 120,000 objects and 30,000,000 individual metrics. Now that’s a lot of data! Now these remote collectors come in different size configurations as well.
- Standard = 2 vCPUs and 4GBs of memory
- Large = 4 vCPUs and 16GBs of memory
In figure 2, this is what a clustered installation would look like and where remote collectors fit in.
Figure 2 – A vROps cluster
As you can see, the main cluster or the master, replica and data nodes all share a database and analytics processing engine, but the remote collector does not. It’s goal is simply to act as a vacuum, to collect and push the metrics collected from those remote data centers back to the main cluster for processing and storage.
All together this makes for a fantastic resource for troubleshooting and metrics data retention for historical data. I will caution that the vROps appliance requires a lot of CPU and memory depending on your environments configuration, and you should be sure to have ample resources supporting it. To get the most from this appliance, I’d also recommend at least one dedicated engineer to vROps, as there is a great deal of information to be had, and much to configure and maintain.
A Final Word
As someone who has been responsible for several large deployments, I can tell you this appliance has come a long way from its former days as VMware vCenter Operations Manager, and the developers dedicated to this platform are hard at work making it even better as it becomes the center of the software defined datacenter universe, within the VMware stack.
There are excellent blogs over at VMware that dives deeper into this appliance and it’s capabilities. For more information visit their site via blogs.vmware.com