Re:DigitalOcean App Platform
by manigandham on Thursday October 08, 2020 @11:51PM
(#24726622)
The vast majority of overages are due to user error. These errors would just be shifted to include quota mistakes, which can incur data or service loss. Usage limits might be softer than monetary limits which are bounded by the time dimension, but can still cause problems since they do not discriminate between good vs bad traffic.
Before you go around calling people lazy, I suggest you put more thought into why creating more options for people who are overwhelmed by options is generally not productive and can cause unintended consequences and expose liability. With some more thought, you'll also realize that AWS is optimized for businesses and, as stated, losing customers or data is much worse than paying a higher bill, which can always be negotiated after the fact.
You know. When I hit the storage limit of my SSD it doesn't wipe my data. It just ceases to store more data. When I rent a server for a fixed price and my service is under a DDOS attack then it will simply cease to work for the duration of the attack. If there is a variable service like lamba that charges per execution then lambda can simply cease to run my jobs.
You can neatly separate time based and usage based charges and set a limit for them separately. It doesn't even need to be a monetary limit, it could be a resource based limit. Every service would be limited to 0GB storage, 0GB RAM, 0 nodes, 0 queries, 0 api calls by default and you set the limit to whatever you want. AWS or Google Cloud could then calculate the maximum possible bill for the limits you have chosen. People will can then set their limits so that a surprise bill won't be significantly above their usual bill.
Your comment is lazy and not very creative. You're just throwing your hands up and pretending there is no other way even though cloud providers have created this situation for their own benefit.