I've been setting up a replacement for our aging (and unlicensed, so garbage) ESXi 4 virtualization platform, and the route we've gone down (due to licensing simplicity) is Server 2008 R2 Datacenter (mix of Core and Full) with Hyper-V roles, Failover Clustering, and SCVMM running to manage the cluster, all of this runs using Cluster Shared Storage from our pair of Overland iSCSI SANs (with target mirroring for redundancy).
Now, due to hardware constraints (we need to eventually recycle our ESXi boxes to be Hyper-V hosts), we had to draft in the aid of our "dev" box (home-brew server based on desktop hardware), and even a normal workstation (just so we had enough capacity to clear one of the ESXi hosts to reconfigure as Hyper-V). So we've got a Core i7 920, in an ASUS P6T SE, with 24GB of RAM, in addition to that we've got a Core 2 Quad (Q6600) in an Abit AWD9-MAX, and 8GB of RAM.
We migrated the VMs on one of the ESXi boxes absolutely fine (across both of the temporary Hyper-V boxes detailed above), and have them in a high availability SCVMM cluster, working like a charm. However today we installed Server 2008 R2 Core, with the Hyper-V role and Failover Clustering onto one of the ESXi boxes (IBM x3650, 2x Xeon E5430, 30GB RAM), and intended on Live Migrating some of the VMs off the "dev" box (mainly so it wasn't overcommitted).
However when trying to Live Migrate, I got the error about processors from different manufacturers, and thus I couldn't live migrate. Obviously they are all Intel platforms, and whilst different architectures, they all support Intel Virtualization Technology, and according to Microsoft documentation for R2 Hyper-V it "should" work, assuming you are either going from Older to Newer architecture, or have the "allow migration to a virtual machine host with a different processor" checked on the VM itself (which we do for all VMs).
So, they ARE from the same manufacturer, all have Intel VT, and we have the checkbox ticked that "should" allow Live Migration from the i7 to the Xeons (it works between the i7 and the C2Q). It's a bit of a sticking point now, as it means that at some point we will have to turn off critical services in order to offline migrate some of the VMs (it's not even as if we need to do storage migration as it's all Cluster Shared Storage), all because of a problem that shouldn't exist according to all documentation?
Is there something I'm missing here?
Now, due to hardware constraints (we need to eventually recycle our ESXi boxes to be Hyper-V hosts), we had to draft in the aid of our "dev" box (home-brew server based on desktop hardware), and even a normal workstation (just so we had enough capacity to clear one of the ESXi hosts to reconfigure as Hyper-V). So we've got a Core i7 920, in an ASUS P6T SE, with 24GB of RAM, in addition to that we've got a Core 2 Quad (Q6600) in an Abit AWD9-MAX, and 8GB of RAM.
We migrated the VMs on one of the ESXi boxes absolutely fine (across both of the temporary Hyper-V boxes detailed above), and have them in a high availability SCVMM cluster, working like a charm. However today we installed Server 2008 R2 Core, with the Hyper-V role and Failover Clustering onto one of the ESXi boxes (IBM x3650, 2x Xeon E5430, 30GB RAM), and intended on Live Migrating some of the VMs off the "dev" box (mainly so it wasn't overcommitted).
However when trying to Live Migrate, I got the error about processors from different manufacturers, and thus I couldn't live migrate. Obviously they are all Intel platforms, and whilst different architectures, they all support Intel Virtualization Technology, and according to Microsoft documentation for R2 Hyper-V it "should" work, assuming you are either going from Older to Newer architecture, or have the "allow migration to a virtual machine host with a different processor" checked on the VM itself (which we do for all VMs).
So, they ARE from the same manufacturer, all have Intel VT, and we have the checkbox ticked that "should" allow Live Migration from the i7 to the Xeons (it works between the i7 and the C2Q). It's a bit of a sticking point now, as it means that at some point we will have to turn off critical services in order to offline migrate some of the VMs (it's not even as if we need to do storage migration as it's all Cluster Shared Storage), all because of a problem that shouldn't exist according to all documentation?
Is there something I'm missing here?
