You’ve just finished your shiny new project and made it available to the world. And they love it! That’s great. Now you’ve got to maintain it until people don’t love it. How many projects can you sustain and when do you cut off the unwashed masses from your sweet, sweet awesomeness?
Slogging through a difficult project is challenging for anyone. A conversation today made me think about my approach and consider if there was a better one. It also motivated me to fix some issues with a Terraform course, so that was positive.
When this whole cloud adoption craze started, sysadmins like me were thrown into a new world. Suddenly we had to contend with the cloud operating model where things were constantly shifting below our feet. Operations were more fast paced, infrastructure as code was a thing, scaling VMs was a thing, and what the hell is an object store? Let’s just say for anyone making the change, you may have encountered a fair amount of technical whiplash.
Then Kubernetes rolled onto the scene and once again the poor beleaguered sysadmin had to learn about containers, pods, mountains of YAML, and even more dynamic environments where a workload could just disappear at any moment. Trying times for those used to the gentle churn of on-premises VMs. And worse yet, K8s wasn’t just running in the cloud, it started infiltrating our cherished data centers. First it started as a side project or a collection of weird VMs that one of you business units asked for. Then VMware decided to integrate it directly into vSphere like it was a first class citizen. The gall!
And now we’ve got edge. Frankly, I think sysadmins have had enough. We don’t want another sea change where we have to deal with distributed computing that is truly distributed in tiny data centers peppered across a metro area. We learned your cloud and your IaC and your Kubernetes. Dammit, we are not learning whatever fresh hell comes with edge.
I suppose the good news is, we don’t have to.
Sure we can create Infrastructure as Code, but does it make more sense to create infrastructure using software? This was an idea put forth by Kris Nova, and I have to admit I find her logic a-pleasing.
Blog post: https://nivenly.com/lib/2021-06-13-ias/
With the explosion of popularity in podcasting, I start to wonder what will happen to us small fries out there. Will we get steamrolled by the Spotify’s of the world, or will we sneak by unnoticed as lumbering giants traverse the pod-o-sphere?