Ent Manager Issue? - Object Count 0 for monitored clusters

[tomgreg]

Hello All,

Trying to figure out if this is an issue with my nworks Enterprise Manger dashboard, or a collector, or something else. Basically, when looking at Object Count in Monitoring Group Status I see two clusters as having 0 Objects Monitored. At first I thought this was a license issue, however after we upgraded our license to cover all sockets in the environment (and have some to spare) I still don't see any Objects Monitored for two of my clusters. Moving these clusters to another collector doesn't resolve the issue, nor does a topology rebuild or load balance. All hosts under the cluster(s) are selected in the Connected VMware Servers tab.



Thanks for any insight or help!

Tom

[agolubnichy]

Hello Tom,

Is it possible for you to remove and re-add in nworks UI the connection to the vCenter that manages the Clusters in question? Please, give it a try.

[Alec King]

Hi Tom,
I would also check permissions in vCenter - does the nworks connection account have access to all objects? We need just read-only permissions, but we must have those permissions from the root vCenter node down through all clusters and hosts.

Do the hosts in those clusters appear in SCOM? Do they have current metrics in performance charts, etc?

Cheers
Alec

[tomgreg]

Hello all,
Thanks for taking time to look at this.

1) I don't believe I can re-add the nworks UI connection because it is in use to monitor the rest of our VMware environment (single vCenter server) and I don't want to impact monitoring of the rest of the systems if I can help it. That said, I'll try it if nothing else seems to be working I would try it but just to be clear do you mean Configuring the OpsMgr agent (from the Enterprise Manager Dashboard in SCOM) on the nworks Enterprise Manager Collector server, or from the individual collector that runs the VI-AP Link? (Or something else entirely?)

2) I double checked permissions, and we have read-only from the top down for my nWorks/Veeam monitor account.

As for hosts in SCOM, yes I do see the ESX hosts that fall under those clusters as actively reporting in SCOM (ESX Host State shows them) and the Cluster State health explorer appears to have relevant data. Performance charts also seem to be there (I can create reports that capture info on the hosts and VM's on these clusters).

Here's a screenshot showing the ESX Host State which is populated with hosts in the UCS Conf. Cluster (hosts 12,13,14,16,17):

[Alec King]

Hi Tom,

Well, it looks like you are getting monitoring data. It could be that the '0 objects' is just an artefact in our UI.
You could also try a restart of the nworks Enterprise Manager service - that should not cause any monitoring outage or impact.
Failing that, remove and add the vCenter in our UI could also fix it. But that could cause un-discovery and then re-discovery of the whole environment in SCOM, which could cause some monitoring downtime.

I'd advise open a case with our support, they will guide you to grab relevant logfiles and we'll figure the best way to address this. Thanks!