You buy into hyperconverged for the promise of simple, scalable performance — and for the most part it delivers. But you’re not just plugging the servers into the wall and walking away. You have interactions with VMware hosts, VMs and datastores to think about, and that’s before you start thinking about load balancing and the database and the actual application. A new vRealize Operations (vROps) management pack brings Nutanix into the picture, allowing you to monitor the hyperconverged side-by-side with the virtual.
Figure 1 — Hyperconverged and virtual in a single view
Dashboards like Nutanix Health Investigation (see figure 1) put VMware and Nutanix in a single screen, linking together, for example, the Nutanix view of a VM with the VMware version. When you have a user complaint that an application is slow, you can look at latency from the VMs, hosts, datastore and Nutanix container all in one place. Now you’re not bringing up vSphere and Prism Central in two windows and rubbernecking back and forth.
The Nutanix management pack digs deep into the vROps feature set, enabling capacity analysis on disk (SSD and HDD), IOPS, throughput, CPU and memory throughout the Nutanix system. Figure 2 shows a 200 day horizon before a Nutanix VM will have inadequate memory. Having wide ranging capacity analysis means you can use vROps projects to plan for future capacity needs, both with current workloads and your future apps.
Figure 2 — Time remaining for the VM
Prism Central is a fantastic tool for managing Nutanix, but sometimes alert management can be a little cumbersome. The management pack simplifies things by allowing you to disable individual alerts and filter to the Nutanix alert level you want to see. There might be a low level alert storm going in Nutanix, but you float blissfully above it in vRealize.
Figure 3 — Only see the alerts you want to see
The full power of vRealize can be actualized when your whole stack is inside it. Beyond Nutanix, that means bringing in Dell EMC and Lenovo servers with management packs for the bare metal side of things, associating F5 nodes and virtual appliances with their VM counterparts and bringing in databases like Oracle and MSSQL.