It’s been an exciting week y'all. Kanye (or “Ye” apparently) still continues his downward spiral of gaffes like in the interview with Lex Fridman (warning, explicit language), while Elon has ruffled the feathers of the biggest little blue bird. October has come and gone and taken with it (in my opinion) the best of the Candy Holidays: Halloween. Sure, Christmas is great with all the presents and songs and nostalgic movie flashbacks (to which we add the Christmas Story Sequel this year), but it also comes with the price of buying all the right gifts, travel issues, family meltdowns, and extreme cold depending on where you celebrate.
Halloween is fun, people assume any number of alternate personas they want (or have the raw materials to create), and then go out in said personas seeking out candy, giving out candy or escorting candy seekers. The only real disappointment this year was the fact I didn't get Doom in any of my full-sized candy bars (because sadly, I didn’t get any full-sized candy bars :( ).
My wife is something of a witch, and because of this, I’m also reminded of the ancient roots of my favorite Candy Holiday in the Celtic holiday Samhain.
Winter is Coming
Samhain is one of the four seasonal Celtic celebrations along with Imbolc, Bealtaine and Lughnasadh (none of which I pronounce correctly) and marks the beginning of winter and the new year in Celtic traditions. And though evidence of burial mound alignment with Samhain is close to 5000 years old, the concepts of mumming and guising, or wearing costumes and entertaining for rewards (mainly food) has been documented for around 500 years. And while I’m sure folks weren’t getting sugar hangovers from “cheap”, bite-sized candy 500 years ago, they absolutely knew of the darker, colder days coming and the fact that it was time to slow down, conserve energy and figure out how to keep warm and fed through winter. Slowing down in times of scarcity is importance for our survival, and a few extra pounds from trick gotten treats will surely be beneficial.
Cloud Winter is Coming
Recently I’ve noticed a lot of slowing down and rethinking in the automation and cloud space. I think this is actually a good thing as it shows some reflection based on some difficulties folks have had around actually moving to new models of application operations. The first major reason for this rethink is due, not surprisingly, to cost.
Many folks were promised that their operations in the cloud would be more efficient for any number of reasons, but, apparently, many companies are still waiting for their cloud Return On Investment. A lot of this has to do with the fact that “lift-n-shift” operations aren’t really the best way to save money in the cloud and moving to more “cloud-specific” SaaS services (not to be confused with “cloud native” which somehow got usurped by the Kubernetes Kamp) requires some deliberate design engineering in order to truly realize cost optimization in the cloud. The folks at the Duckbill Group’s article on Why Cloud Finance Is Broken and Ineffective has some real gems and is rooted in operational experience with a wide variety of organizations trying to get their costs under control.
To build an effective cloud finance team, your organization should:
Rethink what you think you know about cloud spending. Operate from a new assumption: Architecture and costs are the same thing.
Work to better understand how your architecture impacts your costs and how your specific cost drivers behave. The majority of your efforts should be on understanding cost drivers instead of RI/SP management and identifying idle resources.
Build processes into your engineering release cycles for ongoing cost optimization efforts. A little bit of time spent every week will have much better results than a lot of effort every quarter.
Staff your cloud finance efforts with engineering, supported by finance — not the other way around.
Acknowledge that your tools aren’t the complete solution. Tools are not a replacement for people; tools augment people.
The key to fixing broken, ineffective cloud finance teams is realizing that cost management is primarily an engineering problem.
I think this is one of the more important pieces of advice I’ve seen in awhile and it’s reinforced by the number of Cloud 2.0/3.0/4.0 project names I’ve heard while talking with customers. A lot of articles out there like Why do cloud migrations fail? start out with “lack of planning” as the number one reason for cloud migration failures or project overages. While I agree with this, I’d say that design along with planning is equally as important. Remember, “cost management is primarily an engineering problem.”
Further making things difficult are the macroeconomic issues everyone is facing. While all costs are going up all around us, the crunch is also being felt in the cloud sector due to inflation as well.
"We expect public cloud prices in Europe to increase at least 30 percent in 2023, causing an almighty shock to many of their customers who are also trying to get their costs down and leading to a lot of tension and a lot of aggravation between you and them and providers."
And while Mike Norris’ reaction below may be a bit hyperbolic, there is definitely some reason to be concerned.
Mike Norris, CEO at Computacenter, one of Europe’s largest resellers, did not comment on the specter of prices heading northwards, but said the most frequent request from clients was how to get control of cloud costs.
“Cost control is the biggest challenge with the cloud,” he said on stage. “Because it’s not software-as-a-service, it’s software as a hostage. You have no choice, you can’t get out of it.”
Yikes! I’m not sure we’ll see the acronym SaaH popularized anytime soon, but try to reduce lock-in as much as possible.
On the extreme end of slowing down are folks like those at Basecamp who have decided to ditch the cloud altogether.
Renting computers is (mostly) a bad deal for medium-sized companies like ours with stable growth. The savings promised in reduced complexity never materialized. So we’re making our plans to leave.
Their opinion is that there are two really good use cases for cloud and that none of their workloads fit into those categories anymore.
Simple applications that can leverage cloud-specific SaaS tooling
Highly irregular workloads that need to scale up or down on demand
While this may be an extreme reaction to cloud costs, it’s good to see applications of design and planning as well as re-evaluation of performance and cost helping an organization make decisions.
Mr. Norris above may be wrong in that folks can’t leave the cloud, but it’s highly likely the move will be costly in any number of dimensions. To make matters worse, the global supply chain for semiconductors is also under duress, and lead times to secure hardware has increased anywhere from 2 to 3 times as long as organizations experienced pre-pandemic. One clear benefit of operating in the cloud are the economies of scale for hardware acquisitions that you don’t have to manage. If you’re considering maintaining your data centers or moving back in, model hardware acquisition times into your planning and design.
While these stories of impending Cloud Winter may have you ready to hide under a weighted blanket (which my wife loves BTW), I think the most important lesson to take from all of this is that deliberately designing your application infrastructure in the cloud is necessary to reduce the costs of cloud operations. And as you begin to rationalize the tools you will use to build your cloud platforms (while also avoiding cloud vendor lock-in), curl up next to the fire and learn how HashiCorp’s concepts of the Cloud Operating Model can give you the leverage you need to operate efficiently wherever you decide to run your applications.
Aquaman Out.
P.S. For those of you in the Southern Hemisphere like our esteemed colleague Grant Orchard, sorry for the Northern-oriented seasonal rant, if anything, at least you have summer to look forward to.