I have always been interested in virtualization, but I have never had the opportunity or resources to take on such a project. Well, since I’ve been working where I currently do, I have had the resources. It has been a challenge trying to get others to jump on board and really move into virtualization, and honestly I am still working on it. The current problem that we are now faced with is that our SAN is out of date. I am running an MD3000i with one RAID controller, so I just pray that it doesn’t go down.
Anyway, I learned quite a bit from the experience. It started with comparing Hyper-V to VMware, during which I found out a lot of information about both products, and what kind of environment they are best suited for. I ended up going with Hyper-V and using my MD3000i with my Dell Switch.
Everything from installing Hyper-V Server Core to configuring the server went very smoothly. The bumps started happening when I was trying to configure the SAN and validate my cluster. All with causing little downtime to the servers running. The first problem that I had was a scsi 3 persistent reservation, which was fixed by a firmware upgrade to v2.0. Then I had to install MPIO on the core machines, one worked great, the other not so great. Then came in Dell Tech Support, who just took over my computer and started clicking around, and I learned nothing.
Once the storage was configured I ran into the problem of having my Virtual Networks configured differently. On one host they were not allowing Host Access, and on the other they were. The validation wizard did not like such configuration. Once I figured that out, everything else went very well.
I was testing Virtual Machine Manager 2012, but ran into some problems when trying to make the virtual machine highly available. The option was grayed out, and when I tried moving the machine into the CSV it would just fail. So, I decided I would fall back on Virtual Machine Manager 2008, which allows me to move the existing VMs into a CSV and automatically configured high availability.
Then I came into the fact that I needed to put my VMM server into the Failover Cluster so that I can still manage everything if a host goes offline. To do that, I had to power down my VMM server and export it using “Hyper-V Manager” then import it on the other host. Then, I had to add it to the cluster to configure High Availability. Once that was taken care of, everything worked. I’m proud to say that I can live migrate my VMM server, which helps me out a great deal.
So, now I just have to make some decisions on what needs to be highly available and what doesn’t, even in the case of my RAID controller going down.