New on LowEndTalk? Please Register and read our Community Rules.
All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.
All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.
Hello Merchants of Complexity
Interesting take on the cloudmania 
‘Merchants of Complexity’: Why 37Signals Abandoned the Cloud
Provisioning a fleet of servers in 30 minutes is magic, but it is not a requirement that comes up that often for most businesses, David Heinemeier Hansson argues.
More here:
https://thenewstack.io/merchants-of-complexity-why-37signals-abandoned-the-cloud/
Thanked by 1raindog308
Comments
I mean the original argument for cloud-based pricing model was that these services will be available for immediate provisioning meaning you won't have capital tied up until you need the hardware and, by pricing it by the hour, you'll only pay for what you use (assumed to be cheaper in the long-run). Remember these pricing structures were made when unit cost of cloud (compared to today) was significantly higher. Also, with the availability of computing the assumption is that resources are available as you scale up-or-down... in case your website "hits it big" on reddit.
The problem is that the magnitude in cost over time has gone up that the "economic" argument of cloud is no longer true. Especially since the cloud market is already hyper competitive. While it's not a perfect market (winner does not take all), a few select concentrated winners do end up winning.
With added on value such as high availability, anycast, and other "configure-as-you-go" solutions (I'm only talking primarily about the enterprise cloud market leaders, such as AWS, GCP, and Azure and not bare-metal vendors or other market sectors), the unit cost of cloud has gone up to a point that for the top vendors they're no longer competing as much on unit costs but rather a risk management or technology access (why spend hours if not weeks self-engineer these solutions when you can build your infrastructure within a few minutes on Azure?). The idea is that these companies will invest in infrastructure and software resources to let you build your own configuration.
TNS article simply focuses on unit economics from the perspective of small and medium businesses and talks about bringing these responsibilities in-house (meaning you have more control over these solutions). However, that also means they're going to have additional costs with hiring (expertise), management, and often troubleshooting.
For a small and medium businesses, I agree that building everything on the cloud becomes an economic question. On a purely computing-focused cost-benefit analysis, you'll find that AWS/GCP/Azure can't compete with Bare-metal vendors or even medium/small cloud vendors like you see advertised here on LET. But for larger corporations, the value proposition works where cloud makes a reasonable argument.
They way I think of it, for someone like 37Signals I'd suggest something like DigitalOcean or Akamai Cloud (Linode) may have a better value proposition than AWS/Azure/GCP. But even then, you're not going to be able to beat the pricing on colocating your own hardware and bringing all that talent in-house at a certain scale.
There's a reason why many people don't sing praise about AWS/Azure/GCP on forums like these.
Believe it or not I can't go for cloud services because of this issue of not knowing and varying pricing based on usage. I find it much simpler to have my own servers with set limits and pricing. However I only once ever had a million page visit site many moons ago when I was University and the host at the time valiantly handled it - those were the days of cheap hosting that really went beyond the call. I have been tried tireless since then to build something that would attract a million legit visitors so I could learn and build something that would handle it but no dice.
That pretty much sums it up.
That's what LinkedIn found out.
https://www.cnbc.com/2023/12/14/linkedin-shelved-plan-to-migrate-to-microsoft-azure-cloud.html
Every cloud presentation/keynote/etc. I've heard made it clear that you don't go to the cloud to save money if you're a big organization, unless you're in a niche market where constantly ballooning resources is vital (e.g., Netflix) or you must have many POPs all over the world.