VSAN and vCenter Operations Interop

Continuing on my set of posts related to Virtual SAN (VSAN) interoperability, let’s take a look at how vCenter Operations Manager (vC Ops for short) integrates with Virtual SAN. vC Ops version 5.8, which was released in December 2013, recognizes the VSAN datastore and can report various characteristics, as you might expect. Although vC Ops 5.8 was released around 3 months before VSAN GA’ed, this release works with ESXi 5.5U1 and vCenter 5.5U1, the vSphere release which introduced VSAN. However, this release of vC Ops does not present all the ‘storage’ metrics for VSAN like it does for datastores based on other storage types. But, having said that, there are still a number of useful vC Ops views and metrics that you might find useful which this post will cover.

First, there is the relationship view, which shows you the various ESXi hosts and VMs leveraging the VSAN datastore. This is a nice ‘getting started’ view to ensure that everything is still connected.

The second view is of the VSAN datastore itself. Now here, you might typically find to see both capacity and performance information. Well, in this version of vC Ops, it is really just a capacity view only. There are no other metrics displayed. The workload badge is partly reliant on disk I/O, but since we don’t pick that up for the VSAN datastore, workload is never fully calculated.

As you can see, disk space consumption is visible, but none of the IOPS, Throughput or Latency metrics for the datastore are displayed. Remember that a VSAN datastore is not a traditional datastore, presented via a single entity  (LUN or Volume) located on a storage array for example. A VSAN datastore is distributed across multiple hosts and multiple disks in a cluster. This makes is at lot more difficult to come up with a set of ‘unified’ datastore metrics that are meaningful and useful (which I’m pretty sure we will see in future releases). So, at least in this release of vC Ops, there is no datastore metrics for VSAN. But other metrics are still visible. For example, if I select a particular virtual machine, I can still go to the storage section and view I/O metrics related to individual disks.

So how do you examine performance characteristics of a VSAN datastore?

Well, you can still get per CPU and Memory metrics for virtual machines in vC Ops. And per VM statistics and per VMDK statics are also available for virtual machines deployed on the VSAN datastore (these per-VM and per-VMDK metrics are visible in both vC Ops and in the vSphere web client). With this information, you can identify a troublesome VM or even VMDK for instance.

Now that you have identified the VM or VMs that are experiencing performance issues, if you do need to continue troubleshooting VSAN performance issues related to disk and network, you can use the VSAN Observer utility that is part of the Ruby vSphere Console (RVC)  and which ships with vCenter server 5.5U1 (both Windows and Appliance versions). Duncan does a good post here on how to get started with RVC and the VSAN Observer.

Bottom line – vC Ops 5.8 can give you VSAN datastore capacity and connectivity views, as well as per-VM and per-VMDK metrics, and this will give you a good idea of how things are behaving. Additional tools are necessary to get VSAN datastore specific metrics. As I mentioned previously, I’m pretty confident that this is something which will change over time with newer releases of vC Ops.