aws lambda and server api endpoint deployed on railway pricing
The endpoint is a scraper with puppeteer. Just scrapes some data and returns result. Expected to be called 10000 times a month. Every call on average would run for 5s
a) endopint is hosted as lambda on aws
b) endpoint is hosted on railway.app as express server
Can anyone help me to build and intuition how these two cases would compare in pricing?
3 Replies
@kstulgys - in my opinion, (and you may hate me for this answer) is, I would deploy it and wait till the estimate rolls in. Usually after a day, our estimation is correct around 90 percent of the time.
Hey @Angelo no I don't hate you. This gives me an intuition that there is no obvious answer. I would appreciate if I could get a few more feedbacks/opinions on this.
Sure thing, I think there is just too many factors when it comes to being able to give a committed quote. Until we have a system where we can have users commit to usage (and at that point this would be something like an enterprise contract.) we won't be able to have the level of price stability that you might look for. (Unless there is something like deploying and instance. Which wouldn't be "usage based")