Turns out nothing was wrong.. apart from the fact I had not written any data to the test vhdx files... After installing os the test VM was exported imported and initial replication worked perfectly.. who would have guessed
Server 2022 to 2016 Hyper-V initial replication fails - Event 32594 Hyper-v-VMMS Virtual Disks not present on replica VM
We have recently built a new 2 node failover cluster using Server 2022 to replace an ageing 2016 cluster built to the same specification.
The current setup Replicates to another off site 2016 server.
The plan is to set the 2022 server up to replicate to the 2016 server. in order to achieve this the VM configuration version must be set to 8.0 (supported on both 2016 and 2022)
During testing I have created a new VM (using PowerShell to ensure config file is version 8.0 and generation 2 VM)
If I initiate replication without a VM on the replica server and use the option to send initial replication over the network
Everything works and the VM replication is health
However I choose not to do it as all the config files VHDX etc end up all over the place using GUID as folder name etc make troubleshooting problematic.. and you don't want that in the even you actually need to failover..
The preferred approve is to export the vm.. import is in place on the replica and initiate replication using an "Existing VM on the replica server"
This is how I have done it for the last 5 years on the 2016 server so I know it works well and is constant.
When doing this same task of exporting the new VM, importing the VM on the replica server from the 2022 server to the 2016 server (or 2016 to 2022)
The replication succeeds to identify the existing VM and mark it as replica (event Hyper-V-VMMS ID:32500) and setting up replication however fails initial replication with this error on the replica
Hyper-v-vmms event id 32594 The virtual disks for the Replica virtual machine 'TESTVM' are not present on the Replica server (Virtual machine ID XXXXXXXX-XXXX-XXXX-XXXX-XXXXXXXXXXXX). Restore a copy of the virtual machine on the Replica server or choose a different mode for initial replication, and then try again.
I just don't understand why I'm receiving this error. After importing the VM i am able to start it successfully with no issues so clearly it can access the Disks. And this also proves that creating the VM on the 2022 server with version 8.0 does create a compatible VM with 2016.
2x Dell R650 running 2022 Datacentre with shared SAS/DAS for shared storage
2x 10Gbps Intel x710 NIC in a Switch Embedded Team (SET) Note: x710 do not support RDMA) this is the hyper-v Production network swithc.
2x 10Gbps Intel x710 NIC in a NIC team using LACP for cluster traffic.
1 Dell R640 running 2016 Datacenter with SAS/DAS
2x 10Gbps in NIC team for production
2x 10Gbps in NIC team for cluster traffic
Anyone run into this before and know a solution ?
Sign in to comment
Sort by: Most helpful