Recently, I taught my very first hands-on two-day virtual lab. It seemed a natural outgrowth of my speaking at TechMentor and Live360 conferences; the same company sponsored it. The topic was one I knew in some depth, and the pay was good, so why not?
The problems started with the timing and how things change over time. When I was first approached about the class and subsequently agreed to do it, Azure Labs was still a valid option. Unfortunately, Microsoft shut down that option one week before I went to sign up for it, so I had to find a different option. My next plan was to build the lab at home on some spare equipment, but I ran into some technical difficulties there with nested virtualization and the limitations of older hardware. Another idea was to replicate my home environment in the cloud. Another failure since you can’t run nested VMware Workstation on Azure Virtual Desktop.
What I wound up with was a truly hybrid solution. The class was migrating from VMWare to Azure Virtual Desktop, so I built a VMWare Horizon environment locally. Then, I was going to use ReadyTech as a classroom management solution. This worked wonderfully in testing. My problem came when I went to spin up the full class, and I ran into issues with Azure vCPU quotas!
My failure was changing the size of the class VMs and not getting the correct vCPU quota adjusted! There are TWO quotas, one for general vCPUs in a region and one for the specific family of vCPU sizes. You have to increase BOTH to the right size, and it can take as much as 24 hours for this to happen. So, by the time the adjustment was made, getting the images validated and ready was impossible.
I corrected and adapted, and I think the class went pretty well. The lesson learned: Just because you believe everything is lined up, test it again a few days before!
The experience was pretty stressful, but I’m already considering ways to improve my process. I use this continuous improvement process in many areas: try your best, find the weak spots, improve those areas, try again, find other weak spots, and repeat. We can sit around forever trying to make something absolutely perfect and never actually TRY it out, which means we never get anywhere. Conversely, preparing too little or testing too little also leads to disaster. The balance is the key!
I hope to have the opportunity to do this again. Maybe next time, a ONE-day class!