Article number
000003758
Affected Versions
All
Source Hypervisor
All
Target Hypervisor
All

Problem with Empty List of Datastores when Selecting Recovery Host Cluster

Viewed 194 times

Summary

An administrator may attempt to create or edit a VPG where a host cluster is selected to replicate to, however, no datastores can be selected for use within that host cluster.

Root Cause

One of the hosts in the host cluster is offline and unable to retrieve any metadata.

Symptoms

In the Create/Edit VPG wizard under the Replication Tab, after selecting a Recovery Host Cluster, the datastore option is blank

If choosing a specific host, the datastores are available to be selected and appear in the list.

On the recovery side, you will likely have a host that reports: "Failed to collect metadata for host: XXX"

The ZVM logs will show similar events to the below:

,VCenterProxy,ComputeResourceEntityInformationToComputeResourceReflection,Could not find HostConfigInfo for host-XXXin ClusterComputeResource/domain-XXX,

,Basics,GetHostFolderDescendantEntityInformation,Null HostConfigInfo for host host-XXX. There is a known VC bug which requires restarting VC in order to get HostConfigInfo again.,

,HypervisorPotentials,ReplaceEntityNames,"RecoveryComputeResourceEligibility:ID = [Host: host-XXX, server abcd1234-5678-5678-7890-abcd12345678]; EntityName = [HOSTCLUSTER]host_name; Is Eligible = False; Not Eligible Reason = Host = hostnameĀ doesn't have any associated datastores; Reason = NoAssociatedDatastores; Resource Pool ID = was added into potentials",

GUI logs may present something similar to the below after selecting the host cluster in the Zerto GUI:
{
date: "2019-09-05T14:05:29.547Z",
operation: "$exceptionHandler",
protocol: "$provide",
type: "decorator",
content: "message : TypeError: Cannot read property 'InternalDatastoreName' of null
at P (https://zvm_hostname:9669/scripts/scripts-0a74423129-min.js:33:23925)
at https://zvm_hostname:9669/scripts/scripts-0a74423129-min.js:33:24665
at https://zvm_hostname:9669/scripts/package-d89af9662b.min.js:6505:44
at m.$digest (https://zvm_hostname:9669/scripts/package-d89af9662b.min.js:6516:185)
at https://zvm_hostname:9669/scripts/package-d89af9662b.min.js:6519:326
at Pg.completeTask (https://zvm_hostname:9669/scripts/package-d89af9662b.min.js:6531:15)
at https://zvm_hostname:9669/scripts/package-d89af9662b.min.js:6416:489,
cause : Possibly unhandled rejection: {}"
}

NOTE: The same results also occur when editing at the VM level.

Solution

To resolve the issue:

  1. Ensure all datastores are mounted to all hosts in the host cluster.

  2. Power on/reconnect the host that is not communicating within the host cluster.

  3. If the host cannot be powered on/was decommissioned, remove the host from the host cluster.