Devices failing to be added to inventory

[gilesogram]

,We are working on a deployment of FNMS 2023R2 and have the following setup:

• 1 x VM running FNMS 2023R2 on Network A
• 2 x VMs running Beacons (one on Network A and one on Network B)

Each Beacon is configured with a site and associated subnet, and there are looking for VMware vCentre and ESXi hosts (together with the VMs running on those systems).

Discoveries are running on both beacons and the results files as being uploaded to the FNMS. Only the VMware ESXi hosts and vCentres are appearing the Inventory list, and we are seeing no virtual machines under each vCentre/Host.

There are another 151 devices in the discovery list which are NOT appearing the inventory list.

How can we check/investigate the reason for these not appearing as inventoried items?

Also - we have some physical workstations, which are being discovered by the beacons, which aren't being added to the inventory. We have installed an agent on one of the (as a test), but the beacon is doing a zero-touch inventory and that workstations is not showing as having the agent installed.

Thanks in advance

[mfeinman · Accepted solution]

Some things to consider ...

According to your first image, the Gathering Device Inventory step ended with errors.  Have you examined the discovery log to determine what those errors were? (Most often, it's connectivity errors or bad vCenter credentials).

Discovery doesn't result in an agent being deployed unless your discovery rule is set to allow FNMS to adopt the devices it discovers.  (Adoption is simply having the beacon deploy the agent on a discovered device).  How are you getting the agent on the devices? Are you deploying them yourself through "third party deployment" as Flexera calls it?  Discovery just means "hey, I found a device"; it doesn't mean that an agent will be installed on it.

For the single device that has the agent deployed on it, have you checked tracker.log to determine that inventory ran and that it was successfully uploaded to a beacon? I would start there.

Also, you wouldn't see results in the UI until Reconciliation runs.  Is your Reconciliation running properly?

--Mark

[mfeinman · Accepted solution]

Some things to consider ...

According to your first image, the Gathering Device Inventory step ended with errors.  Have you examined the discovery log to determine what those errors were? (Most often, it's connectivity errors or bad vCenter credentials).

Discovery doesn't result in an agent being deployed unless your discovery rule is set to allow FNMS to adopt the devices it discovers.  (Adoption is simply having the beacon deploy the agent on a discovered device).  How are you getting the agent on the devices? Are you deploying them yourself through "third party deployment" as Flexera calls it?  Discovery just means "hey, I found a device"; it doesn't mean that an agent will be installed on it.

For the single device that has the agent deployed on it, have you checked tracker.log to determine that inventory ran and that it was successfully uploaded to a beacon? I would start there.

Also, you wouldn't see results in the UI until Reconciliation runs.  Is your Reconciliation running properly?

--Mark

[gilesogram]

Hello Mark,

Thanks for the update.

I have attached the discovery logs from beacon2 and this shows that the discovery seems to be ok. There are some hosts in the Device Inventory (.49 to .58) reporting as RPC errors, but this is because these are our ESXi hosts not Windows devices.

Re the Agent - we aren't selecting adoption (yet), but we have just install an agent on workstation06 manually as a test. I did wondered how the agent gets it policy and found the following article to run the mgspolicy.exe command manually:

Inventory Agent ERROR: No upload servers are configured (tracker.log) - Community (flexera.com)

Running this on workstation06 to the beacon2 returns a error (which we are trying to also investigate/fix!) - see attached screenshot

[JohnSorensenDK · Accepted solution]

@gilesogram 

You can force the agent to upload to a specific beacon using a parameter when running it manually, i.e., 

ndtrack -t Machine -o UploadLocation=https://<Beacon-FQDN>/ManageSoftRL

Thanks,

[gilesogram]

Many thanks. We have now been able to work out the agent collection on our Linux estate now and seem to have a working FNMS POC solution in place.