Some vmware probe VMs no longer show data in Metric Viewer
Article ID: 245031
Updated On:
Products
Issue/Introduction
We have in our environment several Vmware probes configured with MCS. Everything is working fine except for 2 devices/vm's/hosts that are showing the following behavior:
- These devices do not have any recent QOS in the Database.
- The devices show duplicated entries in the CM_computer_system tables for each device and display NO IP Addresses.
Additional notes:
These virtual machines could not be found in the AC - System read-only trees in the VMWARE probe configuration below the vCenter.
When searching the inventory for the IP of the affected server we can see that the VM correspondent to IP is actually assigned to another VM's name.
Other devices have correct active QOS being stored.
Environment
- Release: DX UIM 20.4.* / 23.4
- Component: Vmware 7.x
Cause
- Some VM's were decommissioned and the IP address was assigned to some other VMs
Resolution
As the devices were removed from the vCenter and IPs were assigned to another device there was a discovery mismatch between these devices.
We shall try to remove them (all devices) from the inventory and let the discovery try to correlate the 'active' devices again.
Steps to follow:
1. Stop the vmware probe
2. Select the Inventory and search for the affected device:
"VM_NAME"
Select it from the list and then go to the 3 dot menu on the right and select "Delete Devices"
3. Deselect Prevent Rediscovery, Acknowledge alarms, and Delete measures
4. Repeat point 2, and 3, for all affected VMs
5. Clear the niscache on the robot where vmware probe is installed: How to clear the niscache on an individual robot (broadcom.com)
6. Start the vmware probe
7. Restart the discovery_server
8. Wait for 15-20 minutes, then check the inventory again to see if the issue is resolved
Feedback
