Case ID: 02138419
We came across an issue with replication between two Hyper-V hosts on different domains:
Additional Details: A replication job between to Hyper-V hosts is failing due to RPC not connecting. RPC connects fine while testing outside of Veeam. Erorr message: 26/04/2017 04:42:26 :: Failed to process replication task Error: The RPC server is unavailable. (Exception from HRESULT: 0x800706BA)
We checked WMI connectivity manually - all fine.
We checked Event Viewer connectivity - all fine.
The exact message in the logs was:
Invoking WMI method: 'AddKvpItems'[26.04.2017 04:41:12] <15> Warning Host is added by ip address or DNS name. Check host availability by netbios name[26.04.2017 04:41:12] <15> Info WMI reconnecting...[26.04.2017 04:41:12] <15> Error Failed to execute WMI Query.[26.04.2017 04:41:12] <15> Error Failed to execute WMI Query.
The resolution was as follows:
Add NETBIOS names of all opposite end servers to HOSTS file on both Hyper-V hosts
i.e. on HOST1:
VH02 22.214.171.124 (Hyper-V host)
VHB02 126.96.36.199 (Backup VM running Veeam)
VHS02 188.8.131.52 (Server being backed up)
VH01 184.108.40.206 (Hyper-V Host)
VHB01 220.127.116.11 (Backup VM running Veeam)
VHS01 18.104.22.168 (Server being backed up)
What would be useful:
a) Better Error Message - this is the feature request - when detecting this error, suggest this as a possibility. You should also be able to easily check NETBIOS availability and report true/false.
b) I'm not an expert on RPC or Hyper-V by any means - however, I'm sure that RPC should be able to work via IP address instead of needing the name? Is there a legacy piece of coding that is attempting to use the VM name? Hosts file is not a very elegant work around. Is there a better fix?