Dell EMC Enterprise Storage Analytics, the plugin to vRealize Operations, is now out with version 6.2. I haven’t posted on ESA since 6.0 because 6.1 did not make any changes for PowerMax. This release, however, has a couple new features.
The major features are:
- FC-NVMe tolerance
- vVol replication support
- (vROps 8.4 support – OK this really isn’t a major feature but I have a lot of customers asking about it)
And then to support the above, changes were required to the following:
- Dashboard items
- Topology items
The FC-NVMe tolerance basically means that having the emulation configured on your array won’t break ESA. We collect metrics on the object itself, but we do not show the relationships beyond the director/port. Here is the view from the PowerMax Topology dashboard. I’ve highlighted the director which is linked to the director ports, though as you can see, to no other objects.
The same metrics we collect for FA directors/ports, are collected for the FN emulation.
It’s unclear if we will expand the FC-NVMe topology relationships up the stack to say devices. NVMeoF has very little adoption in general, but even less so in VMware environments. I’ve previously covered why this is the case. But the good news here is if you use FC-NVMe, ESA will not break. If you do use it and would like more topology, drop me a comment.
The second feature is vVol replication support. VASA replication groups (type is VASA_Async) are built at the vVol Storage Container level because a replication group for vVols is a relationship between these containers. So to see the replication groups for vVols, highlight your vVol storage container in the PowerMax Topology dashboard. In the Health Tree widget on the right, at the very top are all the configured VASA replication groups.
If you drill-down to one of the RDFGs, in this case 35, you can get a summary.
The relationship to the storage container is the depth of the topology.
We do not expose the devices in a VASA replication group as we do for a regular RDFG below:
For the most part, vVol, or VASA replication groups have the same metrics as regular replication groups, though there are some quirks. For example, two metrics which normally have the same value are Device Count under the Configuration category and Devices in Session (count) under the Performance category. In a VASA replication group, though, the Device Count is always 0, while the metric Devices in Session (count) is accurate.
This reflects what you would see (or not see) in Unisphere when viewing the replication group. Here is a regular RDF group with the volume count and a VASA group with no entry.
For those who use vVols, this will come as no surprise since we abstract the vVol objects as you cannot manipulate the SRDF relationship outside of the vCenter.
As always, if you use ESA with PowerMax and you are looking for metrics or objects we don’t have yet, leave me a comment and I’ll get it on the roadmap.