Do you need a #multicloud#strategy? Most likely, not. Just use one cloud that fits your needs.
The chances are that your project is multi-#cloud, anyway. There are nuances, of course.
One is TRAFFIC. The cost can go up depending on the path it follows.
A thread 👇🧵
1/6
The bigger the company, the bigger the chance that you are already using multiple #clouds. Generic public clouds (#AWS, #GCP, #Azure), private clouds (#VMWare, #Nutanix), and specialized clouds (#Cloudflare, #Vercel) can all be combined together for the benefit of your org.
2/6
If you have just a few GBs of traffic a month, then this would most likely not have a big impact on your #cloud#costs. But if you have more than a few GBs per hour, a #multicloud setup may bring you some surprises.
3/6
Traffic in/out of your load balancers, #CDNs, on-prem resources, and between clouds is not free.
For example in #AWS, transferring one gigabyte into or out of their networks may cost you 0.01 - 0.08💲per GB.
This may result in 💵hundreds or even 💵💵 thousands of dollars paid for high traffic configurations (e.g. image processing, financial transactions, analytical data streams).
Even within the same cloud, traffic may be costly if it spans several regions or availability zones
5/6
So, after all, if you decide to go with multiple clouds or multi-regional setups, then you definitely need a multi-cloud strategy that should take traffic costs into account 🙂.
6/6
• • •
Missing some Tweet in this thread? You can try to
force a refresh
If you are still having a "Monoliths vs microservices" debate, do you include the following aspects in a discussion?
👇🧵
1/9
💡 Cognitive load on the team. How difficult it is to understand and change the code?
💡 Deployment and platform maturity. How time-consuming is it to build, deploy, and observe?
💡 Scalability and operability of the final setup. Can we handle traffic growth or spikes?
2/9
I've seen teams that are pretty happy with a monolith. One code base - easy to search, easy to reason about, easy to refactor, easy to deploy. Codebase size is ~100-200K LOC. Cognitive load is small.
3/9