Long story short, I've been working with support on this, and it blows me a way that they're not able to provide a suitable solution. I don't think I'm asking too much here out of a monitoring solution. I'm writing here as a last-ditch effort to see if any of the wizards here are able to do what support can't.
Some background:
My team manages the virtual infrastructure only housed in the datacenters. We do not support servers in the field. Therefore, we only want to be alerted on virtual servers in the datacenter. This consists of 9 clusters in VMware between our 3 DCs. the 10th cluster is what the field team supports.
I want all virtual servers to be managed nodes nodes. VMAN pulls in VMs, and an AD network discovery manages all servers from AD. I want this to be an automatic process. I want any VM to be a managed node in Solarwinds without having to manually manage it. This basically automates our on-boarding/off-boarding of VMs with less human intervention. This piece appears to work.
I want some sort of identifier on all of the managed VMs that reside in those 9 clusters. I want to only receive alerts on the 1600 or so VMs in those clusters, and I don't care about anything else. I thought I figured this out, but apparently it caused some issues where some VMs were not being alerted, even though they were found in the scope of the alert.
My alert:
When working with support, they found the issue below. This is due to building an alert on "Virtual Machine" instead of "Volume". Myself and support have not found any way to get Volume to see the virtual clusters under the scope. All I want is for the alert to be applied to the VMs in those clusters. Even custom properties or groups don't make what I want possible it seems.
Issue reported by support:
Upon further checking, the reason it was not captured because the Volume Summary Free Space Percentage condition in alert has different calculations if you compare to the volume Percent Available from the actual Volumes.
Tables that I used to check the details.
Volumes (for actual volume percent available)
VIM_VirtualMachinesVolume
VIM_VirtualMachines
VM_A
NodeId: 3960
C:\ : 2372792320 kb
VirtualMachineID: 13695
VirtualMachineID MountPoint Capacity FreeSpace
13695 E:\ 1.60924E+11 64462110720
13695 C:\ 63897071616 2374778880
VolumeSummaryFreeSpace: 66828341248 (Bytes)
This is not captured from Alert because the free space above 10% when you convert from KB to GB and get the percentage
VM_B
NodeId: 3634
C:\ : 0 kb
VirtualMachineID: 14216
VirtualMachineID MountPoint Capacity FreeSpace
14216 C:\ 42842714112 16625664
VolumeSummaryFreeSpace: 16625664 (Bytes)
This is captured from Alert because the free space lower from 10% when you convert from KB to GB and get the percentage
Their workaround:
Need to find proper way to alert on individual Virtual Machine component volume space utilization in VMAN Orion (SolarWinds Support)
I'm going to give the workaround a shot, but I'd still like to figure out how to only alert on managed nodes in those clusters without having to manually go to each node and set a commend or property. I'd like to find a way to make it so all VMs under those clusters get marked as the Data Center team's servers, and we can easy apply that scope to all alerts we need to create for our VMs.
Is anyone doing what I'm trying to accomplish? If so, how? I want to put in the work now so that there's far less room for human error down the road when on-boarding or decom'ing VMs in our environment.
Thank you!
