So, it’s been a little while since I’ve written the previous installment in this series regarding our new hyperconverged infrastructure project. There are several reasons for this. I was away at a conference; I took some time off; I had a few other projects to complete; and I have been somewhat busy actually doing the work of migrating VM’s to our new Scale Computing platform.
First of all, a confession: I cheated. Well, ok, it’s not really cheating since we own the old and the new systems.
What I mean is, Scale Computing provides an organized schedule of training and implementation help with their premium installation services. I decided not to wait for all of the training and unlike Luke Skywalker, I did not find my long-lost father nor did I put the rebel alliance in jeopardy. I did, however, migrate VMs. I simply used the documentation provided by Scale to import the VMs into the new HCI.
Our existing environment is Hyper-V. The export/import process is very straightforward: Start by creating a template VM in Scale and export it. This becomes the basis for the new VM. Now, export a VM from Hyper-V. Replace the exported virtual disk in the Scale Computing template with the disk file from the exported Hyper-V VM.. Now import the modified template VM back into the Scale Cluster. Other than the need to reconfigure your NIC and load a few device drivers, that’s it. Scale’s KB article provides all the details.
So far I’ve moved 12 VMs this way. Late last week I completed the next step in my Jedi training and learned how to use the Move utility that Scale offers. It’s a variation on the old Doubletake program (now owned by Carbonite) if you’re familiar with that. Relatively simple and painless, it allows you to migrate a VM by synchronizing it into a template VM in the cluster. This is a much preferred method for larger VMs since the export of a Hyper-V VM and the subsequent import back into Scale can be rather time-consuming. Whereas swapping a VM that’s been synchronized through the Move utility takes just a few minutes.
So far there are two incredibly impressive things about Scale Computing. First and perhaps most importantly, it just works. Everything that we were told it would do, it does, in a simple and error-free way. Secondly, every person that I have worked with at Scale is extremely knowledgeable about their platform and eager to share their knowledge and best practices. They are incredibly helpful and friendly. I realize that helpful and friendly may sound a little cliché, however these attributes are very important at a time like this when we’re moving around the firm’s live data and production systems. It inspires confidence. And perhaps that is Scale Computing’s number one product: Confidence.
Next time: the final chapter. We didn’t use The Force, but these are the droids that you were looking for.
Check out the previous posts in this series:
Our Move to Hyperconvergence - Part 1
Our Move to Hyperconvergence - Part 2