Hacker News Reader

Comment by mrkurt | Hacker News Reader

We actually launched with that feature: https://news.ycombinator.com/item?id=22616857

No one took us up on it. What we found is that the majority of people want their stuff to stay up, and the right UX for "shut it down so you don't get billed" is not obvious.

We ended up implementing prepayment instead. If you sign up and buy $25 in credit, we'll just suspend your apps when the credit runs out.

Bandwidth is weird because we have to pay for it (as does every other provider). We aren't yet in a position where we can just make it free without limits. Maybe next year. :)

View on HN

I'm actually very curious: why is bandwidth so much cheaper on more traditional VPS or dedicated server hosts like Hetzner ? This extends to their somewhat new-ish cloud product, where you get 20TB traffic included - even on a tiny instance. And it's 1 Euro per TB after that. [1]

Do they just decide to not profit from bandwidth or are they doing something special that allows them to be so cheap?

[1] https://docs.hetzner.com/robot/general/traffic/

There are three ways to manage bandwidth prices:

1. Put servers where bandwidth is cheap (not Sydney, for example)

2. Constrain throughput per server

3. Buy from cheap transit providers like Cogent

Hetzner does all three. Bandwidth in the US/EU is very cheap. They meter total throughput on their services. And they use cheap providers. None of these are bad choices, just different than ours.

Our product has multiple layers, too. When you connect to a Fly app, you hit our edge, then traffic goes to a VM that's probably in another region. When you hit a hetzner server, there are no intermediate hops.

We usually pay that three times as data moves from customer VMs to our edges to end users (out from our edge to worker vm, out from worker vm to our edge, out from our edge to end user). Or 10x, in some cases, if data moves from Virginia to Chennai to an end user.

We pay $0.005/GB in the US and $0.9/GB in Chennai. You can see how this might add up. :)

The chennai pricing in incredibly steep. It's weird that bandwidth prices for end-users in India is among the cheapest (if not the cheapest) in the world but enterprise customers are paying so much.

Jio and friends are (were?) subsidizing connectivity and trying to drive growth to make it up later/elsewhere like payments and services.

Service providers come to the opposite problem. Local infra sucks, the market is full of incumbents, and india is generally protective of those markets.

Very similar dynamics with china as well.

$0.9 is a typo, I meant $0.09/GB in Chennai. We're not taking that much of a haircut.

You forgot play the peering game which can lead to substantially cheaper bandwidth when you connect to enough Tier1/2 ISPs. I think Hetzner does this some as well.

I would love to hear the answer to this as well.

My guess is that they either don't profit from bandwidth or they have peering agreements from back when the internet was young.

transit pricing can still be had readily for ~.10 cents per megabit. that's per second mind you. so you pay for capacity not for bits transferred. {$cloudprovider} makes a sweet margin by upselling that to you in terms of bytes transferred. CF wrote a post explaining it [1]

Obviously, you would need to manage servers,colo,network and keep it up on your own, or pay for it. And cloud providers offer alot of value as well. but if you are at the right scale you can diy it (in-house ops/network team) you can save ALOT of money.

1 - https://blog.cloudflare.com/aws-egregious-egress/

Having a billing limit is important to me as I don't want to risk unbounded costs for personal projects.

> we'll just suspend your apps when the credit runs out.

This sounds great! I've looked at Fly.io before but didn't realise this was a thing so didn't go past looking. I'll definitely give Fly.io a test run now. :)

I can't speak for your whole market, but I know for me the pre-loading flow sounds really clunky because I'd have to go and manually add funds each month (right?)

It's understandable if your usage data showed the fee-capping feature just wasn't popular enough to be worth maintaining, though that would surprise me based on this thread (but possibly HN just isn't representative of the whole market)

I’d be happy with a “refill if under $10 up to $x/month”. If they can cut off when you’re out of credit they can presumably do it for other criteria, too.

Do you also have a monthly prepayment subscription where the balance is topped up to a fixed (ideally, optionally calculated automatically via web UI by the user)? Naively that's what I'd expect the solution to look like.

Reply on news.ycombinator.com