We're currently running VMAN 6.1.1 and we've integrated our instance with our NPM instance. Ta-da -- magic datastore data appears for hosts and VMs and -- oh, wait a minute? Why doesn't that host have storage related data? We have 8 or 9 vCenters that we poll data from via VMAN and all are managed in NPM. As a data geek I have a health distrust of data collected until I can be sure that it is reliably collected and reported -- and even then I want to be able to prove that point at any given time.
I have two challenges -- one is VMware hosts that are not managed nodes in NPM and two, and perhaps the more challenging, is VMware hosts that are not reporting storage metrics. I'll share my query for the first one if someone would be so kind as to help with the query for the latter.
Hosts That Are Not Managed in NPM
SELECT [HostID]
,[NodeID]
,VIM_DataCenters.Name AS DataCenter
,VIM_Clusters.Name AS Cluster
,[HostName]
,[PollingSource]
,[BootTime]
FROM [SolarWindsOrion].[dbo].[VIM_Hosts] WITH (NOLOCK)
RIGHT JOIN VIM_DataCenters WITH (NOLOCK) ON VIM_DataCenters.DataCenterID = VIM_Hosts.DataCenterID
RIGHT JOIN VIM_Clusters WITH (NOLOCK) ON VIM_CLusters.ClusterID = VIM_Hosts.ClusterID
WHERE NodeID IS NULL
Run this query and then take the resulting hostIDs and plug them into this URL - [Your_Orion_ServerName]/Orion/VIM/HostDetails.aspx?NetObject=VH:[HostID] This will take you to the Host page and you should see something like the capture below. Click on Manage This Node and follow the prompts.
NOTE: If the node is already managed in NPM you will get a warning that looks like the warning below. If the node can be polled via SNMP simply click cancel and then click next again from Add Node page again. Follow the prompts and you will find that the custom properties are already populated as per the node in NPM. Finish the process and the HostID will disappear from the query above.
If the host is ICMP only you will have to either convert it to an SNMP node in NPM or delete the node in NPM and add it via the process above. I'm not sure why the same process doesn't work for ICMP only nodes but it doesn't.
