Configuring vSphere HA - Cannot install the vCenter Server agent service: "Unknown installer error"
Article ID: 382465
Updated On:
Products
Issue/Introduction
When configure HA for the cluster vCenter shows the error.
Trying to uninstall the existing VIB from the host from SSH , by using ' esxcli software vib remove -n vmware-fdm ' , failed with
The associated volume is a VVFS , it is showing invalid symbolic link as following when doing ' ls -lah /vmfs/volumes/ '
VMkernel.log:
2024-11-14T18:29:57.944Z cpu15:2683524)WARNING: Vol3: 718: Disallowing mounting VFFS volume vffs-xxxx.xxxx02b to host/xxx which is not the one that forma$
2024-11-14T18:29:57.944Z cpu15:2683524)Vol3: 4462: Failed to get object 28 type 1 uuid xxxxx FD 0 gen 0 :Not supported
2024-11-14T18:29:57.944Z cpu15:2683524)WARNING: Fil3: 1539: Failed to reserve volume f533 28 1 xxxxxx xxxxxx 2500a5e4xxxxxx xxxxxxx 0 0 0 0 0 0 0
2024-11-14T18:29:57.944Z cpu15:2683524)Vol3: 4462: Failed to get object 28 type 2 uuid xxxxxx-xxxxxx-axxxx-xxxxxxxx FD 4 gen 1 :Not supported
Environment
VMware vCenter Server
VMware vSphere ESXi 8.0
VMware vSphere ESXi 7.0
Cause
VFFS (Virtual Flash File System) is a derivative of VMFS. It is created when setup virtual flash resource ,i.e. when Host cache is configured for ESXi host. When formatting VFFS volume , ESXi will use the MAC address of the physical NIC as part of the volume UUID (last 12 digits).
By VFFS design, to mount VFFS volume, ESXi will use the current Host UUID(which has the current physical NIC MAC address info) to compare the existing VFFS volume UUID(the last 12 digits). If they are different, ESXi will not mount the associated VFFS.
The current Host UUID could be changed if physical NIC is replaced (as it will have a new MAC address).
Resolution
To resolve the issue :
a. Repaired or deleted the invalid symbolic relate-datastore. And reconfigure vSphere HA.
b. In vSphere Client , follow the steps to remove the Virtual Flash Resource.
c. If the "Virtual Flash Resource Management" does not showing any current configured device ,
1. Identify the device name associated with the VFFS.
2. In vSphere Client , Storage Devices, Pick the device and 'ERASE PARTITIONS' on the device.*
(*The VFFS will have VMFS partition type, but showing 'Not consumed' under 'Datastore column')
3. Remount VMFS from SSH section (' vmkfstools -V ') or Rescan Storage (check both options in the scan list) from vSphere Client.
4. Confirm the invalid symbolic link no longer showing up (from SSH ' ls -lah /vmfs/volume ' )
Re-configure HA on the cluster level, The ESXi host's existing vmware-fdm vib (if there is any) will be removed and new vmware-fdm vib will be installed.
Additional Information
The invalid datastore may as one of vSphere HA's available heartbeat datastore.
Feedback
