vRealize Operations 6.5 became generally available this month and in that release we announced improved scalability for new and existing vR Ops clusters. These improvements bring a new cluster node size, supported memory scale-up options and increased capacity for existing nodes. In this blog post we will cover these product enhancements in further detail. We will also take a quick look at the updated sizing guide.
Monitor Larger Environments in the Same Footprint
Without doing anything, other than upgrading from vR Ops 6.4 to vR Ops 6.5, clusters will gain a 25% increase in the supported number of resources. As you can see from the table below, for Large size nodes, this is a significant boost to the capability to monitor more objects in your environment without increasing the node size or cluster size.
As you can see, the Large node capacity increased by 30,000 objects with the vR Ops 6.5 release. Later in this post we will look at the new sizing guide for details on improved scalability in all nodes.
Simply Scale Up
As your environment grows, it’s now easy to add vR Ops capacity by adding memory to the cluster nodes. The node memory on each node can be scaled up to two times, which allows increased cluster capacity without adding additional nodes. Many customers find that new management packs get released that would add value to their enterprise monitoring. Since this means additional objects will be monitored, the cluster must be adjusted to compensate for the increase in metrics. Increasing the node memory to handle these additional metrics is an easy way to expand a previously deployed vR Ops cluster.
Please keep in mind that all cluster nodes must all be configured equally. So, when scaling up memory plan to upgrade all of the nodes to achieve the improved scalability.
Also, no additional CPU is required. Of course, if you scale your nodes to the next-sized node memory configuration, you are practically changing the node size. In those cases, the vCPU count should be increased to match the node size configuration.
New Node Size – Extra Large
A new node size has been added with the release of vR Ops 6.5 – the Extra Large (or XL) node. This node size, as you can see below, brings much higher capacity per node than was previously available. Customers in very large enterprise environments can deploy up to four XL nodes to achieve cluster capacity that is capable of monitoring up 120,000 objects with smaller cluster sizes.
Why is this important? Consider an environment with 120,000 objects to monitor. Without XL nodes the cluster size would be 12 Large nodes (using vR Ops 6.5 sizing) and those nodes would require a total footprint of 192 vCPUs and 576GB of RAM. Apply the XL node to the same scenario, and the cluster size is 4 nodes with a total resource footprint of 96 vCPUs and 512GB of RAM. Clearly, a more efficient use of resources.
Updated Sizing Spreadsheet
The sizing guide has been updated to support the improved scalability of vR Ops 6.5, including the node capacity improvements and the new XL node size. Highlighted below are the updated and new sections of the Scaling Table.
The sizing guide and spreadsheet can be found at KB 2148829 and you should review the sizing guide when installing or upgrading your cluster. The sizing spreadsheet now also includes tabs for the Management Pack for Storage Devices, so you can plan appropriately if you would like to deploy that management pack as well.
The scalability improvements in vR Ops 6.5 provide customers with increased deployment options and better scale option for currently deployed clusters.