COVID, Supply Chain, and cloud
ALL VIEWS ARE MINE AND MINE ALONE. THEY DO NOT REPRESENT THE VIEWS OF MY EMPLOYER OR ANOTHER COMPANY I AM AFFLIATED WITH.
It goes without saying the lessons learned from COVID as it relates to technology infrastructure. However, with another new variant and a seemingly new resurgence, hopefully the lessons learned have not been quickly forgotten.
The other issue facing technology, now the source of often sarcastic comments, is supply chain issues. Getting the equipment to run computing infrastructure is increasingly difficult. There have been cases where an order is placed, and the hardware is projected 6-9 months out. By that point, the business imperative necessitating the purchase will almost assuredly have evolved or might have gone away.
So, what is to be done? This is where the power of the cloud model comes in. If you know me, you will know that I don’t mean to immediately spin it up in Azure, AWS, GCP, etc. Even some of those public cloud providers are struggling with supply chain issues. Some regions may not be able to provide a particular instance type you want or might have processors from older generations because they too aren’t getting hardware delivered in a timely manner. For those like me thinking it, yes even the cloud has hardware requirements. The public cloud providers also must deal with scaling and usage projections just like a private company would if it were hosting. They have the luxury of purchasing power that most private companies don’t.
We are stuck, yes? Private companies are struggling to get their infrastructure needs, so are the public cloud providers. How can the issue be addressed then?
True cloud computing allows us to blur the lines between hosting infrastructures. While software may depend on hardware to run, it doesn’t need to depend on particular hardware to run. If software is written in such a way that it doesn’t tie itself to the platform it is running on, then new options become available. If your company doesn’t have the necessary resources currently, run it in your GCP tenant. If GCP doesn’t have the service you need in the region you need, look at Azure or AWS. Being able to host things flexibly gives more power back to the organization. Choice is a major bargaining chip. Cisco rep comes in and gives you a great deal, but you must wait 6 months for hardware, but Dell rep can get you hardware in 3 months though it’s a little more expensive. Maybe now you don’t have to make that choice and you can decide to deploy into AWS first, determine true adoption, and then make decisions with live data rather than projections. Now you also have an operating model where if the 6-month promise extends to a 9-month hope, you are still working. And if by chance AWS costs go up, then in 9 months you can pull it back onto some nice shiny new servers in your data center.
The true power of cloud is flexibility. Nobody can predict what the future holds. We may be facing a recession, countries may invade countries that we get supplies from, transportation of goods may improve or worsen…we just don’t know. Understanding what you don’t know, or can’t know, can help you focus on what you do know. There is a particular business imperative that requires a technology to run a process. Can it be deployed at a moment’s notice to any hosting? If not, can it be made to do so. Certainly, there will be things that will inevitably be impacted, but working to minimize that impact is where many should spend time. If the contention for scarce resources can be reduced in one area by allowing to be hosted any way imaginable, then those resources that can’t be hosted that way inherit some flexibility as a benefit.
Cloud operations is a powerful tool. Many companies are struggling with how to make it work but those companies have an advantage. They are preparing themselves now with hard decisions when they have time to figure it out. That’s a much better position to be in then trying to figure it out on the fly.