Published: February 21, 2020
When it comes to cloud providers, it has become a love-hate relationship for me. Which leads me to believe that cloud providers like Linode, DigitalOcean or Heroku is here to stay. Without the need to worry about overrunning your cost and having a specialist to help you build an AWS architecture. When you start a service unexpectedly due to your lack of understanding in using their service like EC2.
These smaller companies have a healthy mix of different type of documentation like videos or step by step guide to making it easy for you to digest and start using within mins.
Due to the emphasis of digital transformation, reducing the cost of ownership and making the IT infrastructure agile to be adaptable to change to the business environment.
A lot of the companies is moving their legacy systems to the cloud or adopting a hybrid approach. By having a combination of both IT and cloud infrastructure to help them scale their IT operations & software development efforts.
By embracing cloud providers for startups, it makes them agile and faster. Especially many incubator or accelerators have a partnership with major cloud providers like Amazon with generous cloud credits for their portfolio companies.
When it comes to the big 3 in cloud providers, you won't go wrong in picking either AWS, Google and Azure. For me, I always hear great things about AWS due to its focus on the developer & startup community.
Therefore adopting a cloud provider like Amazon has been a goto choice for startups in scaling up their operations or just to get started because of their generous credits.
They help startups and developer who are considering in using their service by offering generous credits to allow new startups to scale their operations across the globe. With tools and services to cater to every need of various processes within the software development, data science, web development and many more.
When it comes to Google, what rings the bell the most for me is their focus on data science-related cloud services. Which anyone could use a service to train their machine learning models. Instead of purchasing and maintain a costly data science infrastructure with GPUs to run their machine learning models. Unless it makes sense for some to build one from scratch with deep pockets.
The strong point of having azure cloud service is either your organisation is undergoing digital transformation. Which the IT department is to migrate part of their IT infrastructure within the Microsoft ecosystem to the cloud using Azure. It could also be that you are looking to do anything related to data science projects for your organisation.
The draw for it is that either you starting a small project or they are just starting to test out their idea with limited capital to get started with expected cost per month.
Linode has been the old kid in the block that started in 2003. They had increased on their investment to upgrade it with latest hardware & setup data centres across the globe.
It was pain-free to go through their setup process to get a server up and running. Many Python YouTubers or podcast that i subscribe to. Love using it due to their rock-solid support and ease of deploying Django or Flask applications using Linode.
DigitalOcean is a new kid in the block for the cloud space. They had been building their reputation and traction by focusing on their product and services along with marketing efforts like paying a developer to write technical documentation or articles for them.
I remembered there were times when I need to solve a problem, especially on server deployment for Django. I read an easy to read documentation with A step by step tutorial in getting what I need with DigitalOcean.
I started using Heroku during my first few years as a Django developer with fond memories of using it. Due to their command-line tool and ease to understand guides in deploying my Django web application using Heroku. It continues to be one of my goto choices besides Linode for deploying Django or Flask web applications.
I hope this basic introduction to various cloud providers could help you to pick the cloud provider for your startup. If you are bootstrapping your startup. I would suggest that you could start with just subscription-based cloud providers like Linode or DigitalOcean without access to the credits by major cloud providers.
Once you are ready to scale and you plan to use more of the various services provided with the cloud provider's ecosystem. Which does not require a significant focus on controlling the cost of usage of credits for these cloud providers? By breaking some eggs while you learn to use their various services to the best of your organisation's use case and technical teams ability?