Fly

Infrastructure

load balancing and edge routing

SHOW HN2017

112points
68comments
#402of 2,242 in 2017
Top 18%of 2017 by points

How HN took it

MixedNo clear lean

6 positive, 11 neutral or asking, 0 mixed, 12 critical, out of 29 comments about the project. Thread score -0.21.

Pricing: 1 for, 8 againstPrivacy: 1 for, 1 againstIdea: 2 for, 0 againstPerformance: 0 for, 2 againstDesign: 0 for, 1 against

The thread, comment by comment

Each row is a top-level comment and its replies, in HN's order. Select a cell to open that comment on HN.

positive neutral or mixed critical about another commenter or something else the maker

What HN said

What people liked

  • Interesting product that makes middleware at the edge useful

    From 2 comments
    • This is very cool. I suspect though, you'll eventually be competing with CDNs that are adding API management features. You're starting off with middleware that's smarter than a…
      tyingq on HN
    • This lets you do smart obvious things like route /blog on your home domain to your blog provider. There's a lot of stuff you can do that makes your website better and faster…
      garry on HN
  • Beta user finds it very cool

    From 1 comment
    • I've been running Fly in beta for one of my apps and it's pretty damn cool. Adding a couple more today.
      thomcrowe on HN

Objections

  • Pricing per request is far too expensive, especially at volume

    From 3 comments
    • Wow, this is _really_ expensive! I'm sorry, but $0.05/1000 requests is really really premium. 1000 request/s is $0.05/s. 86400s/day * 30 days * $0.05 -> $129.6K/30d. 1k req/s is…
      Xorlev on HN
    • I'd still say you need to take a hard look at your pricing. Lets assume we're talking about a pretty standard Rails site on Heroku without ridiculous traffic. Lets say 1500…
      Xorlev on HN
    • The pricing structure has similar problems at the low end. A small (2.6mm pageviews per month is tiny) SaaS would presumably have more than one request per page they want to…
      pvg on HN
  • Routing all traffic through a for-profit company raises abuse and trust concerns

    From 1 comment
    • I'm sure the tech is really neat, but after the last CloudFlare debacle, piping all your HTTP requests through a for profit company just seems like it's ripe for abuse. This is…
      secstate on HN
  • SSH tunnel transport limits bulk throughput

    From 1 comment
    • > When your visitors explore your application, their HTTPS requests arrive at the closest regional edge node, buff up with Middleware, then intelligently route to your…
      otterley on HN

Requests

  • Volume-based pricing or tiers

    From 2 comments
    • Some sort of slider or something like CDN's have with scaling number of requests a day and the associated pricing would help. So like 10K a day, 100K a day, 250K etc.
      dubcanada on HN
    • I'd still say you need to take a hard look at your pricing. Lets assume we're talking about a pretty standard Rails site on Heroku without ridiculous traffic. Lets say 1500…
      Xorlev on HN
  • Add API gateway features like rate limits and caching

    From 1 comment
    • This is very cool. I suspect though, you'll eventually be competing with CDNs that are adding API management features. You're starting off with middleware that's smarter than a…
      tyingq on HN

Questions

  • Where are the edge locations and how is routing done

    From 2 comments
    • I just tried this, and the IP for my test site seems to resolve to a Chicago based ISP called Secure Central Network - is that the closest one to UK? Do you have a list of edge…
      marcosscriven on HN
    • That's an anycasted IP. We use one pool of IPs globally and let BGP select the shortest route. If you run a traceroute you'll probably find that you hit a server in Amsterdam.
      mrkurt on HN
  • How is it implemented on Kubernetes

    From 2 comments
    • Looks like a very interesting product. I can see how off-loading middleware can really make teams more productive. Can you go into more detail on how you are running this on k8s?…
      moondev on HN
    • We're actually only running some of it in k8s, the edge nodes are ansible managed bare metal. We do however, work really really well with k8s apps. Our apps are all k8s…
      mrkurt on HN

Alternatives named

  • Cloudflare

    From 2 comments
    • I'm sure the tech is really neat, but after the last CloudFlare debacle, piping all your HTTP requests through a for profit company just seems like it's ripe for abuse. This is…
      secstate on HN
    • Pretty interesting - but doing some number crunching, Fly looks to be two orders of magnitude more expensive than enterprise cloudflare. Which does offer smart routing, load…
      jhgg on HN
  • Fastly

    From 2 comments
    • Looking at the response headers from your own site, everything is being served up from Fastly edge servers. Why wouldn't you serve your own site from your own infrastructure? The…
      andrewbarba on HN
    • If anyone is considering this, consider fastly and custom VCLs too.
      willejs on HN
  • Nginx load balancers in AWS regions

    From 1 comment
    • Wow, this is _really_ expensive! I'm sorry, but $0.05/1000 requests is really really premium. 1000 request/s is $0.05/s. 86400s/day * 30 days * $0.05 -> $129.6K/30d. 1k req/s is…
      Xorlev on HN
  • Digital Ocean Load Balancers

    From 1 comment

Category, keywords, summaries and labels on this page were made by AI (Thread Reader v2) and not checked by hand. Each point links to the comments it came from.

Add a badge to your READMEShown on HackerShow
[![Shown on HackerShow](https://hackershow.org/badge.svg)](https://hackershow.org/show/13985940)