Ask HN: Best option to host an MVP within budget? | Hacker News Reader

Ask HN: Best option to host an MVP within budget?

3 pointstiuPapa7 comments

I have built an MVP, it's a social network, with specific features that target a niche group of user (atm, I just want to see if this works and then make it more general in case it is actually useful the way I think it is).

Problem is I have no experience in hosting and I am not sure what are the things I should be aware of when choosing any particular option.

View on HN

If you just want An Answer, then use Heroku. Throw it up there and see if there's traction before you worry about hosting. Once you find traction, there's a lot more to consider, but for now you're talking about minimal traffic and you can deal with this later.

At that point, you start getting into breakdowns of the various providers out there, and--most importantly--what they offer you in terms of failure resilience. If you know nothing else about software deployment, know that everything can fail (whether in a data center or in AWS or whatever) and that correctly and quickly responding to those failure cases is the core of a modern operations/platform effort.

At the point where a product works and has traction, I'll usually evaluate AWS or GCE (and tend to go with AWS because the tooling, publicly available knowledge, and my personal experience lean that way), because I like having easy answers for "but what if it breaks?" and "how do I protect myself?" in a way that neither less featureful/lower-quality services like either DigitalOcean or a bare-metal provider like OVH can offer. But people with different budgets and risk tolerances can reasonably disagree.

Well, besides an answer I am also looking to learn. The opinion I hear about Heroku, and also my primary concern, is it can become quite expensive to run, especially since I have to use two dynos - one for backend API and one for serving the frontend app. Compared to that will starting with AWS or GCE free tier be a better option? I am ready to deal with a complexity if that means I can squeeze a bit more value out of stuff since budget is a bit too tight atm.

If it gets expensive, you have a Good Problem on your hands.

I don't think trying to learn AWS's stuff should gate you on seeing if your product gets traction. What's stopping you from launching on something simple and easy like Heroku while you learn AWS or GCE?

Ah yeah I didn't think it that way. Thanks.

No problem. Feel free to email me directly if you'd like; this is a large part of my business.

Thanks, I will keep that in mind. :)

for me - either firebase or AWS. depends on what I am trying to build. Good luck!

Reply on news.ycombinator.com