For as predatory as the pricing is, as an outsider I have to keep wondering why anyone uses AWS instead of various alternatives. fly.io, Digital Ocean, even Azure or whatever.
and there always an AWS product for something you need, that you can connect to your existing AWS tools
Tools like S3, EC2, API Gateway, ECS, IAM, Route53, have so much synergy when you use them together. And you can start with only “simple”/“edge” ones, like S3 or EC2.
In the past, I’ve worked at jobs where AWS already has the corp card, so you don’t need to get any new sign offs to use it. That’s a big motivator to keep all jobs under one roof.
In what way? I’ve never used S3, but I’ve used Azure Storage and it did everything that I needed it to. In particular, for the backup tool that I was working on (and still am intermittently when I have the time and spare brain outside of work), I wanted to guarantee that existing backups were immutable even if the machine being backed up is compromised. Azure Append Blobs with a SAS token let my create a capability that allows the backup agent to create new blobs and append to existing ones but not modify any existing ones. It’s not obvious how to do this with S3 and answers on StackOverflow suggest that it isn’t possible.
I think I’ve read a couple of such scenarios where user’s functions get DDOSed and is handed a huge bill. They get in contact with AWS and they forgive the bill. Does this still happen, or they stopped paying heed?
For as predatory as the pricing is, as an outsider I have to keep wondering why anyone uses AWS instead of various alternatives. fly.io, Digital Ocean, even Azure or whatever.
I’d say it boils down to 3 things :
Tools like S3, EC2, API Gateway, ECS, IAM, Route53, have so much synergy when you use them together. And you can start with only “simple”/“edge” ones, like S3 or EC2.
S3 is hard to beat.
In the past, I’ve worked at jobs where AWS already has the corp card, so you don’t need to get any new sign offs to use it. That’s a big motivator to keep all jobs under one roof.
In what way? I’ve never used S3, but I’ve used Azure Storage and it did everything that I needed it to. In particular, for the backup tool that I was working on (and still am intermittently when I have the time and spare brain outside of work), I wanted to guarantee that existing backups were immutable even if the machine being backed up is compromised. Azure Append Blobs with a SAS token let my create a capability that allows the backup agent to create new blobs and append to existing ones but not modify any existing ones. It’s not obvious how to do this with S3 and answers on StackOverflow suggest that it isn’t possible.
I think I’ve read a couple of such scenarios where user’s functions get DDOSed and is handed a huge bill. They get in contact with AWS and they forgive the bill. Does this still happen, or they stopped paying heed?