makes infrastructure easy for developers


185 points | by penberg 11 days ago


  • lbriner 11 days ago
    Sadly, we have all seen these promises, "X makes Y much easier" but you cannot make complex things easy without removing lots of functionality.

    Kubernetes for the basics is actually pretty easy despite what people say, I got a fairly simple cluster running with little pain doesn't take long before you want multiple clusters, or vlan peering, or customising the DNS or.... and that is when it becomes complex.

    What will do? Probably what everyone else does, starts simple, becomes popular and then caves in to the deluge of feature requests until they end up with an Azure/AWS/GCP clone. If it stays simple, lots of people will say that you will outgrow it quickly and need something else, if you increase functionality, you lose your USP of making infrastructure easy.

    I think perhaps the abstractions are the problem, if you are abstracting at the same level as everyone else e.g. docker images, orchestration etc. then I don't understand how it can ever work differently.

    To make my point, the very first comment below (above?) is about container format, a really fundamental thing that noobs are not likely to know about, they will just immediately have some kind of error.

    • mrkurt 11 days ago
      You definitely can't make complex things simple just by removing features.

      What we did, instead, was built low level primitives, then built opinionated PaaS-like magic on top of those.

      If you're running a Phoenix app, `fly launch` gets you going, then `fly deploy` gets you updated.

      If you want to skip the PaaS layer and do something more intense, you can use our Machines API (or use Terraform to power it) and run basically anything you want:

      We are very, very different than k8s. In some ways, we're lower level with more powerful primitives.

      We probably won't build an AWS clone. I don't think devs want that. I also don't think devs want a constrained PaaS that only works for specific kinds of apps.

      I think want devs want is an easy way to deploy a boring apps right now, and underlying infrastructure they can use to solve bigger problems over time.

      • knicholes 11 days ago
        I'm a dev. That's what I want. I want something publicly visible that I can get running in the first sprint so I can show everyone what my idea is looking like as it progresses. I want setting up the DB and certificates to be easy. I want a static IP to point my domain to. I want it all to just work if I have the source code and I tell it to deploy. I want secrets to be included so I don't have to stand up Vault or something. I don't want to set up my own Jenkins. I don't want to deal with resizing volumes in k8s. I don't want to dick with networking rules/configurations that allow my frontend to talk to my backend and my backend to my DB.

        I also don't want to set up my own log aggregator, grafana, and prometheus/alert manager, but for a quick "show everyone your app", I don't need those. I can add that harder crap later when the app shows promise and I actually need to debug performance.

        • aprdm 11 days ago
          So you want a 5 dollars digital ocean, a DNS domain that costs you 10 bucks a year, and a docker-compose up command running nginx/whatever behind it.

          You can scp the tarball of your docker and docker-compose down/up

          If you want even easier, just put your app on docker and run nginx natively via systemd or wahtever.

          • dathinab 11 days ago
            no because that doesn't just work and involves additional work. A lot of additional work.

            Because it mustn't "just work" and then when the project grows you setup something new. It must "just work" and then scale to continue working for the initial and maybe all phases of the software product.

            So e.g. security matters, so you rally don't want to maintain an OS. Sure a OCI image has OS like properties, but by using stateless approaches and carefully select a very thin base image and rebuild it automatically with the newest version of the image you can sidestep most of that work load.

            Then you need auto scaling.

            Then you need to be able to add services on the fly without problems (e.g. another DB) and they need to have their own scaling.

            Then there is networking, as you don't create a super complex applications you don't need anything complex but still something solid.

   does provide all that

            So not only can you get something running trivially, you can the incrementally improve on it and even ship it in many cases without needing to re-design your deployment.

            Looks tbh. pretty grate for small and mid sized companies (or autonomous project of such size in larger companies).

            • aprdm 11 days ago
              You need auto scaling to evaluate an idea works?! That's the post I replied to. You need none of that. Have been running apps in a decent scale for decades and an NGINX+postgres in a single VM will take you to many customers before it becomes a bottle neck
              • dathinab 9 days ago
                > You need auto scaling to evaluate an idea works?!

                But you don't "just" show that something works.

                Once you did show that you are expected to make a MVP and then a MVP which works in production and then add features etc.

                With fly you can just use the same basic setup to go through all of this steps reducing a ton of friction.

                I have seen startups fail because of this friction. I also have seen internal proposals being abandoned because of this friction.

                • aprdm 9 days ago
                  What is less friction than an rsync command to deploy ?

                  I think something happened with a generation of programmers that completely missed the basics and went straight to abstractions.

                  They know react but not JavaScript, they know k8s but not linux… and they barely know how a computer works. They read a lot of stuff online in blogs but no books. They think moving fast is what some company marketing material says it is.

                  Look anything you build on top of the basics will be more complex than the basics. It’s worth it to abstract complexity when that arises, when you don’t have it, it’s not necessary and only causes you to move slowly.

                  Nothing will be faster to iterate than a file on disk if you’re working alone and want to validate an idea. You can literally rsync the folder you’re dev-ing from on a git tag and change a symlink in a single command. That will take you very far with no friction.

                  What I’ve seen failing is people who find the non existing problems they have more interesting than delivering value

          • knicholes 11 days ago
            $5/mo will host my webapp and a postgres db? The domain has to happen either way, but when it's an idea I'm playing with, I don't want to pay $5/mo as my projects move pretty slowly due to prioritization of full-time work and family.

            But if I'm going to do all that, I might as well just install k8s and deploy with the cli. I've already gone through the learning curve. But I don't want to do that. I don't want other people that might join my project to have to learn k8s for a tiny side project.

            • aprdm 11 days ago
              lol, apps that are probably way higher scale than your side project have been running with a db on the same host for decades now. And that jump to k8s is a big jump.
            • ipaddr 11 days ago
              You can host both on the same machine.

              Get a free Oracle cloud account if on a budget.

        • intelVISA 11 days ago
          You can get a cheap VPS For ~$3pcm just about anywhere
        • farmin 11 days ago
          You want firebase?
          • knicholes 11 days ago
            Okay, fine. That list clearly wasn't sufficiently exhaustive. I want to use the technologies I want to use and not be locked in to certain technologies.
            • farmin 11 days ago
              Gotcha. I'm just testing Firebase now so was front mind. I come from a python, sql, setup own VPS on Ubuntu etc kind of background. Cloud functions and nosql is all quite different.
    • chrisabrams 11 days ago
      > What will do? Probably what everyone else does, starts simple, becomes popular and then caves

      No, mrkurt will not cave, I can guarantee you that. Fly will be a platform that says no to feature requests that don't make sense for their customer base.

      I have no affiliation with Fly, other than I've used it on and off since the beginning of the platform's existence. They're a veteran team that knows how to build platforms. I definitely trust them to go in the right direction with their roadmap, and all my new projects go on Fly.

      • pier25 11 days ago
        > No, mrkurt will not cave, I can guarantee you that.

        I concur.

        Fly is not your typical startup with dreams of becoming the next big corp monster.

        They are just a bunch of talented people with a vision having fun making cool stuff.

        • capableweb 11 days ago
          Not to break your rose-tinted glasses, I think is pretty cool too. But it's a for-profit company with outside investment, coming from investors who expect a return on their investment, one way or another, and they likely have some influence on how the company will act.

          Same has been said for every company who taken outside investment ever. "But no, Heroku/Figma/GitHub/X are different, they really do care about their users and would never sell/go public/Y", and then a couple of years later we end up in the same position.

          It might not even be up to the "bunch of talented people" in the end, what they have to do to survive or to grow. But grow they have to, unless investors are fine with getting their ROI over 10-50 years rather than 1-10 years. A growing usually comes with some pain.

          • mrkurt 11 days ago
            This is a 10-50 year company. You're not wrong, though, we are either building something big or we're building something that will fail. Our bet is that a developer first public cloud is important and needs to exist.

            Which means, there will be one of three outcomes:

            1. We are correct, and manage to build the right thing. We'll get to work on this forever.

            2. We are correct, but not the right group to build it. We fail.

            3. We are incorrect, and the world doesn't need a public cloud for devs. We fail, and I become a carpenter.

            We have the same incentives as our investors. That doesn't mean it'll work. It does mean that we all believe that we're building a product for developers.

            We're pretty good at surviving, so far. And there are early signs that we're good at growing. There's reason to be hopeful. :)

            • morelisp 11 days ago
              I adore and hope for 1), but

              > We have the same incentives as our investors.

              this is wrong, even if today it looks right because the different incentives result in the same concrete things.

              You have a company; your goal is to make the company succeed. Investors have a portfolio; their goal is to make the portfolio succeed. Your company succeeding is only one aspect of their portfolio succeeding, and one whose importance and externalities can change drastically for reasons outside your control.

              • mrkurt 11 days ago
                This is all true. But my goal is to make the portfolio succeed.
            • KrishnaShripad 11 days ago
              Only issue stopping me from using Fly: bandwidth costs. I am spoiled by DigitalOcean's 1TB tier per droplet created (along with bandwidth pooling if you have more droplets).
              • gsanderson 11 days ago
                That's true but Fly's PaaS is perhaps more comparable to DigitalOcean's App Platform (its PaaS). For that, DO include 100GB for free while Fly include 160GB. Beyond that, DO charge $0.10/GB while Fly charge $0.02-$0.12/GB. Cheaper bandwidth for Fly's Machines (more comparable to droplets, as far as I can see) would be neat though.
                • KrishnaShripad 11 days ago
                  Yes I agree that Fly's PaaS is more comparable to DO's App Platform and definitely has better price model. However, I don't use DO's App Platform for exactly the same reasons as why I don't use Fly's PaaS. I like services which provide bandwidth for cheap (or even free — Cloudflare Images and Cloudflare R2 comes to mind). Ideally, I want to be charged only for features I use and not for egress which I have no control over. But I know that is not feasible from business point of view (unless you are Cloudflare).
                  • tptacek 11 days ago
                    It's OK that we're not a good fit for every user; in fact, you should be suspicious of providers that claim otherwise.
            • capableweb 11 days ago
              > 1. We are correct, and manage to build the right thing. We'll get to work on this forever.

              Maybe I'm too pessimistic, I apologize if that's the case. But I fail to see how the company could ever work on "the right thing" "forever" since there is outside investment in the company. Do these investor not want a return on their investment at one point? If it's in 1 year or 10, eventually they are gonna want you to either go public, or get bought by another company, both of which makes the mission goal change from "the right thing" to "the profitable thing" at that moment.

              But again, maybe I'm just overly pessimistic based on bad experiences with VC funded companies.

              • mrkurt 11 days ago
                Oh when I say "the right thing", it includes actually making money. We're not just building a dev focused cloud because it's fun (it is fun!). We think it's a good business with a very large market. An IPO would be fine!
                • mbesto 11 days ago
                  > An IPO would be fine!

                  And now instead of venture capitalists you have Wall Street to answer to.

                  • tptacek 11 days ago
                    (Taking notes) If we IPO, we lose this message board argument. Got it!
                    • mbesto 8 days ago
                      Your snarky response actually proves my point. The responses from the individuals (including the CEO) are meaningless without action. The only action (beyond writing words) the company has undertaken is to take money from venture capitalists.
            • sangnoir 11 days ago
              I've modified the first option

              1.We are correct, and manage to build the right thing [developer first cloud]. We'll get to work on this *until we exhaust that market and are forced to grow beyond it"

              • mrkurt 11 days ago
                That's a pretty big market. Heroku got to ~$500mm of revenue while they funneled their most valuable customers to AWS. Keeping those customers around (by, say, giving them lower level primitives) would have 10x'ed that, I bet.
                • sangnoir 11 days ago
                  That's a great point. I hope it's big enough for you to not reach its boundaries, because transitioning to a larger market will fundamentally alter the nature of the org. Speaking as customer of organizations that made the transition, and as someone who lived through one on the inside
      • Thaxll 11 days ago
        You should look at the Multi-region PostgreSQL feature they did, not their best work imo.

        • simonw 11 days ago
          That fly-replay header is legit one of the smartest pieces of infrastructure engineering designs I've encountered in years.
        • mrkurt 11 days ago
          Hey! I love this feature. What don't you like?
        • xmonkee 11 days ago
          I was thinking about adopting this feature in the near future. Would love to know your reservations against it.
      • Dma54rhs 11 days ago
        Have other people invested money in them? If that's the case, sooner or later they don't call the shots, but rather who owns the capital and wants it to grow.
        • mrkurt 11 days ago
          I think if you look at the companies you like, you'll notice that it's not the investors who call the shots. There are a handful of very large, very profitable, developer focused companies that investors love because they remain developer focused.

          There are also companies that never figured it out because "developer focused" is not the right business model for them. Those are, I think, the companies that make us all feel burned.

          Heroku is one of those companies where "developer focus" is not the right business model. Salesforce has a model, it's working very well, Heroku's doesn't fit.

    • ignoramous 11 days ago
      > it doesn't take long before you want multiple clusters, or vlan peering, or customising the DNS

      Then, Fly is not for such an application. Just not yet. I mean, we wouldn't buy a snowboard and complain we couldn't go skiing. Different tools.

      The point really is, for a similar thing that which Fly is capable of (and other NewCloud services to an extent like,,,,,,,, etc), you're better off NOT using AWS/GCP/Azure. I certainly have found it to be true for whatever toy apps I build.

    • paulgb 11 days ago
      > Sadly, we have all seen these promises, "X makes Y much easier" but you cannot make complex things easy without removing lots of functionality.

      There's certainly a limit, but it makes me so sad that developers see the current state of orchestration and say “welp, it's a complex problem, guess this is as good as it gets” (not you specifically, but it's a common sentiment on HN.)

      Sure, there will always be use cases that require getting down to a lower level, but there's definitely space for reducing complexity for common use cases.

    • diceduckmonk 11 days ago
      > "X makes Y much easier" but you cannot make complex things easy without removing lots of functionality.

      The question is whether customers need or want that "functionality".

    • tootie 11 days ago
      Yeah it's neat but idk who the target market for this is. I can run a custom-designed site with all sorts of features using or Shopify or Hubspot. Once I need a real platform and backend services, you'll need a team that can spin up Docker images in AWS or the like and the full gamut of DIY platform tools. Platforms like appeal to semi-mature orgs but they will either die or outgrow it pretty quickly.
    • ngrilly 11 days ago
      It sounds like you’re commenting in general terms without having looked at what is actually doing. Yes, choosing the right abstractions is the problem, and what makes really interesting is that they chose different ones. It’s really well explained in their docs and blog posts.
      • ipaddr 11 days ago
        Having to read through docs and blog posts to understand what unique abstractions doesn't help those asking. I've read through docs, blog posts, etc. It's for developers who don't want to setup and maintain an environment. Other providers offer this. You end up paying far more than a five dollar droplet in most cases and a lot more than the free Oracle tier offers but much less compared to big cloud like AWS. They offer a free tier and a community helping onboarding. The goal is ecosystem lock-in and they provide enough to win over a certain group in the middle. The fear is the freebies given today will be paid for by the lock-in effects tomorrow.
        • tptacek 11 days ago
          We host Docker containers. As you aptly observe: you can boot up a Docker container just about anywhere. I'm not seeing the lock-in.
          • ipaddr 11 days ago
            It's the ecosystem things like Upstash Redis Database, edgeDB
            • tptacek 11 days ago
              That's Redis, Redis, Postgres, and a database we don't do anything with, for those following along with this.

              Unless "Upstash Redis Database" is one thing? In which case, that's "Redis, and a database we don't do anything with".

              I'm still not seeing the lock-in.

              Whatever EdgeDB is: we're happy if you run it here! If it has like, a company running it as a service, we're very happy to talk to them, too. But we're going out of our way to avoid weird custom services, like Amazon's 19 different messaging services, that have platform-specific APIs. Our APIs are Docker, IPv6, DNS, Redis, and Postgres. Our Postgres is literally just an app running on; you can run it yourself if you like --- or, I guess, run it somewhere else.

  • mroche 11 days ago
    Spent last night migrating a Discord bot from Heroku to Fly as a result of the upcoming closure of the Heroku free dynos. Overall fairly painless, though I opted to provide my own images rather than use a source-to-image pipeline and discovered one little quirk: images require a Docker V2 formatted manifest. I use Buildah and Podman in my workflows which default to the OCI format. This was simple enough to solve once I figured it out, but I only found one or two forum posts on it and spent a lot of time trying to figure out why the deployment couldn't find an image that I manually pushed to
    • bongobingo1 11 days ago
      For others, you can adjust the format with `podman build --format=docker` (or set `BUILDAH_FORMAT=docker`) or possibly you can push it separately?

      Wonder why they don't use OCI format.

      • mroche 11 days ago
        Yeah, I use Buildah for creating my images, so the environment variable or specifying commit --format docker would work. Before I did that I just re-pushed the image itself using the v2s2 format, so:

            # podman inspect | jq ".[].ManifestType"
            # podman push --format v2s2
  • TurningCanadian 11 days ago
    There are a bunch of rough edges.

    Your server doesn't have a static IP for outgoing requests, so to use it with RDS, you can't just open up a port on the RDS side. (They want you to set up your own proxy)

    The NFS kernel module isn't installed, so you can't use EFS. (They suggest some 3rd party userland tool)

    They expect you to set up VPN access with Wireguard for any connections to your containers. You can't just SCP your files to a volume. It's so much more hassle than connecting with kubectl cp or scp, especially if you're hoping to script things.

    All that said, I'm happy to see competition in the "we'll run your docker image" space.

    • tptacek 11 days ago
      This all seems fair. Just a note: we don't generally expect our users to "set up" WireGuard --- we bake it into `flyctl`, our CLI, which automatically brings up WireGuard sessions as needed. A month or so ago I merged a PR to add sftp support to our SSH server; sftp probably works now with "native" WireGuard, and when I get unburied I'll work out an sftp client interface for `flyctl` itself.

      It's a rough edge, to be sure! I just wouldn't want to leave it as "we think the status quo is the right way to handle getting files to and from instances".

  • arkadiyt 11 days ago
    I like Fly but can't migrate my apps to them because they don't support cron jobs - they've been promising it for more than a year now:
    • albertgoeswoof 11 days ago
      I run cron jobs on fly, using a process. Eg with a rails app running Puma:

          web = "bundle exec puma -C config/puma.rb"
          worker = "config/"
      And the shell script looks like this:

        printenv | grep -v "no_proxy" >> /etc/environment
        cron -f
      Although I would say that cron isn’t a great solution for containerized apps on most platforms, it seems like scheduled processes need a rethink for todays infra.
      • tonyhb 11 days ago
        We're building serverless scheduled processes, queues, and event-driven systems at It's early days, and we agree — they've needed a refresh to adapt to modern practices.
    • shortdiv 11 days ago
      We just released scheduled machines that will help with this. It's similar to the scheduler add-on in heroku and supports general intervals (hourly, daily, weekly, monthly). Feature announcement here ->
  • pier25 11 days ago
    I've been using Fly for 2 years now. Overall I'm very happy and I recommend it to everyone.

    If you just want to run a container in multiple regions with anycast, Fly is really the best option out there IMHO. Nothing comes close.

    There are some rough edges for certain use cases but they keep polishing the service and the DX keeps getting better.

    Personally the only features I'm missing today are:

    - PG backups/snapshots. AFAIK these are coming in the form of virtual disk snapshots.

    - Scale apps from zero to say 100 VMs like Cloud Run does. There's some autoscaling right now and the machines API, but still needs more polish. Specially for certain use cases like concurrent CPU heavy tasks (video encoding, etc). AFAIK some form of this is also coming in the next months.

  • aaronbrethorst 11 days ago
    "Although everybody talks about how Kubernetes, for example, is complex and hard, most companies out there use it"

    I am very skeptical of this claim.

    • tyingq 11 days ago
      Me too, though "Most of the Fortune 500" or similar is probably true. It tends to find it's way in somewhere if the org is large enough.
  • mrcwinn 11 days ago
    I absolutely love Fly. Cannot recommend them enough for the right use cases.
    • Kydlaw 11 days ago
      Could someone elaborate on the right "use cases"?

      Just curious about other's people definition, as I imagined as a more self-served BaaS (mostly for web applications)

      • janee 11 days ago
        My guess would be something like:

        - startups, small teams, no devops.

        - horizontally scalable monoliths.

        - no complex infra needs.

        Haven't used but strongly considering migrating over from Heroku and above is roughly what I would say is a good fit for Heroku so hopefully it's the same on

        • jtokoph 11 days ago
          I haven’t done it myself yet (soon), but they seem great for even multi service apps. All apps in an org are automatically connected to each other via a wireguard network. Internal communication and service discovery is provided by special internal dns resolution.

          It feels like the simplicity of Heroku deployments with the power and security of an AWS VPC.

      • kasey_junk 11 days ago
        The big win is web applications that are read heavy and have geographically dispersed users.
  • tschuehly 11 days ago
    The thing is about all these tools is that starting out with a simple VPS is much more cost effective? has 4 cores for 5€ and you can scale up to 10 cores and 60 gb ram for 27€ a month. And if you need dedicated cores 12 cores with 96GB is 130€ vs 8 dedicated cores with 64 GB ram for 550€

    Choose a docker image and just docker-compose up your application.

    If you outgrow that, you might aswell switch to kubernetes and aws/gcp/azure

    • kasey_junk 11 days ago
      If you don’t need edge compute (which you do if you have customers dispersed geographically) then what you say is true.

      But if you do no amount of Kubernetes on the old school cloud providers is going to get you there. You will encounter the hard problems fly solves for.

      • akrymski 11 days ago
        Do most web apps need edge compute? Don't most backends talk to a central DB server anyway?

        Unless data is geographically sharded as well, is there really a benefit? Collaborative apps perhaps?

        • kasey_junk 11 days ago
          Read replica caches at the edge are pretty standard (whether the main db or a cache layer like redis).

          I think the killer app on fly is actually a geo aware sql db such as cockroach. That as a managed offering puts fly above and beyond anything we’ve had before.

          I suspect they know this.

          • akrymski 10 days ago
            Caching at the edge could be done by an API gateway or nginx.

            Geo aware SQL DB sounds like a lot of added complexity. What is the latency trade off in practice? 100ms ping time is probably small compared to query execution time, especially if your backend returns everything the frontend needs in one response.

            I understand someone at the scale of Amazon wanting to shave ms off page loads. But most web apps?

            • kasey_junk 10 days ago
              Well you have to run an api gateway or nginx at the edge…

              The last web app I worked on had a latency budget of 200ms. The one before that it was 400.

              Cross pacific queries can take 200 of that off the top. Even transcontinental is close to 80. With query times you can at least work to improve them.

              Now it’s not as simple as just locality but that’s a starting point.

    • netcyrax 11 days ago
      Is this contabo reliable? Their prices are extremely low (way more than Hetzner).
    • zinclozenge 11 days ago
      How are they offering such low prices? Overprovisioning users?
      • dangerbird2 11 days ago
        One possibility is that they have fewer regions than something like AWS, so they can put their data centers somewhere where they get favorable electricity/cooling costs.
  • shafyy 11 days ago
    Has somebody used Render and Fly and can comment on the differences? I switched from Herokku to Render, and really like the good UX and docs. Never tried Fly though.
    • mrkurt 11 days ago
      Render is very good. People seem to prefer us (Fly) if they like using a CLI, and they prefer render if they'd rather have a full in browser experience.

      If you're trying all the things, is also a good option.

      • samwillis 11 days ago
        I’m currently looking to move off Heroku, leaning towards Fly but Render and Railway look compelling too. However I’m baffled by Railways pricing, the estimating tool on their pricing page is in fractional units and seems to suggest a “resources used” pricing model. The documentation is completely lacking in what this means. Are they autoscaling constantly, or do I have to manage that. How can a customer have a website, on a container service, where they use 0.3cpu per month but be fully available at all times? My assumption was that I need at least 1cpu/month for a constantly running container, unless that scale too zero with super quick startup. Can someone explain it to me?

        (I think I read Fly is planning to add scale to zero to their normal service, they currently have it at the api level with “fly machines”)

        The other thing I want is a completely hands off managed DB with point in time restore. None of those three have that yet. Crunchy Data looks perfect but are not “in cloud” with them, only being on AWS/Azure/GCP. If one of those three added that capability in house I would probably just go for it.

        • leodriesch 11 days ago
          My understanding of the 0.x CPU measurement is that you're sharing a single CPU core with other running applications.
      • rurban 10 days ago
        After Heroku, I've tried koyeb, railway and fly.

        koyeb was too hard to setup for me. railway easier, but the images were extremely unstable.

        fly was easiest to setup, esp. DNS for custom domains and let's encrypt, and works fine with docker images. there's no GitHub app, but the docs for a deploy action were good enough.

      • zoomzoom 11 days ago
        It's amazing how many really interesting and deeply thought-out dev platforms are being built right now. We're lucky! Aside from these 3, replit is doing unique things by integrating the development experience from writing code all the way through hosting, and while they don't compete with a full PaaS like fly for hosting complex apps, for smaller services they are better than people realize already.

        I would also be remiss not to mention Coherence ( [I'm a cofounder] where we're trying to deliver some of the same magic as the best in class PaaS's above, but we are running your workloads in your own AWS or GCP account. We're really excited about the potential future of a great developer experience that can be delivered as a service instead of rebuilt over and again by platform teams in-house.

      • shafyy 10 days ago
        Thanks, makes sense!
  • Daegalus 11 days ago
    I tried Fly recently for the first time to run a few tiny servers for the challenges on

    It has bee pretty great and painless. I have my own docker servers, and so on, but I don't have local registry setup, so dealing with getting images moved around, and all that hassle was going to be annoying.

    Made a free account, did the `flyctl deploy` after making the config file and it just worked moments later. Really a nice flow.

    Not sure yet if I would use it for anything else, but this was nice and easy so its definitely on my list of things to check for new projects.

  • odiroot 10 days ago
    I move my private projects from Heroku to It wasn't as easy as advertised, but oh well, it works now.

    There's one thing that's straight out frustrating: no support for easy environment variables management. Yes you can add secrets but it's hard to read them back. Not everything is a secret, e.g. log level.

    • tptacek 10 days ago
      You can set environment variables directly in your `fly.toml`.

      You can't read secrets back once you've set them. That's because we can't read secrets back once we've set them --- at least, our API can't. The API has write-only access to our secret storage.

      • odiroot 8 days ago
        I'm fully aware of this but it doesn't help. I'd have to commit every change to variables into Git.

        Environment is not code. And Heroku had it figured out.

        And mind you, is trying to sell itself as natural Heroku competitor.

  • diceduckmonk 11 days ago
    It seems originally focused on Elixir, but has identified the product-market need for app server + Postgres.

    We didn't know about when we chose GCP for this setup.

    The initial setup on GCP was exceedingly painful. We used CloudRun for our app server, with the value prop being that "it just works". It didn't. Our container failed to start with zero logs from our servers. Stackdriver was of no help. Eventually we found a Stackoverflow thread revealing that CloudRun didn't like Docker images built from Macs. As always, GCP's official docs and resources are incoherent. GCP docs address a hundred things you don't care about, and the signal-to-noise percentage is in the low teens, if we're being generous. We had to chase down half a dozen bureaucratic things to get our CloudRun app to see and talk to CloudSQL. Apparently with, you just run a command to provision Postgres, and pass in an environment variable to your app.

    We consoled ourselves that GCP was difficult to setup but now it's set-and-forget. This is also a lie. This week we saw elevated and unexplained 5xx. First was CloudRun randomly disconnecting from CloudSQL. As AWS measures reliability in terms of 9s, the way GCP DevRel responded to this bug is that this is a distributed system and therefore acceptable that things just fail a reliably human-reproducable 1%+ error rate. Yesterday we saw botnet traffic scanning for vulnerabilities on our app. This happens if you're on the web, not inherently GCP's fault. We have GCP's Cloud Load balancer setup but it's not very smart. We were able to manually block specific IP addresses but it's no where as meaningful as Cloudflare. Not a fan of Cloudflare the company but their products address a need. The botnet somehow knocked over our "Serverless VPC connection" to CloudSQL. Basically what that is is a proxy server that you are forced to setup because CloudRun can't actually talk to cloudSQL. All the auto-scaling claims of GCP's serverless are diminished if we are forced to introduce a single point of failures like this in the loop. That serverless VPC connection requires a minimum of 2+ VMs, so the scale-to-zero of CloudRun is no more.

    Our experience with GCP is constantly having to come up with workarounds and addressing their accidental complexity. This should not be the customer's problem. For example, CloudSQL doesn't have an interface to query your databases. If you use a private IP for security, you can't even use GCP's command line tools to access this. We found out that GCE VMs are automatically networked to talk to CloudSQL. We ended up creating a "bastion" GCE VM instance and setup Postgres CLI tools in order to do ad-hoc queries of our DB state. For this, we just needed to the cheapest VM but GCP makes even this difficult. As for Stackdriver, it's still been an annoyingly painful UI.

    • jon-wood 11 days ago
      If I remember correctly Fly weren't focused on Elixir at first, and instead were just trying to be a better Heroku, but then one of their developers realised Elixir would be a good fit for one of the components they were writing, and off the back of that realised "Heroku, but with a private network to allow comms between nodes" was a really good fit for Elixir applications.
      • tptacek 11 days ago
        Yes, the platform came before the Elixir love; it just happened that Elixir is the development ecosystem that best takes advantage of a distributed platform with private networking, and that Phoenix LiveView really screams when you can run it within 10s of milliseconds of all your users.
  • vasilakisfil 11 days ago
    I skimmed through their docs but couldn't find whether they:

    * support for abstract apps, not only HTTP/web apps like heroku, let's say I want to deploy a SIP app

    * support for HTTP/2 and potentially HTTP/3

    If they do support these two I would say it's enough to be considered Heroku killer.

    • mrkurt 11 days ago
      You can route arbitrary TCP or UDP services through our load balancing layer just fine. I'm not sure what SIP actually needs, but it might work. We don't currently have a way to route TCP connections directly to individual VMs, so stuff like WebRTC doesn't work.

      We do support HTTP/2.

  • turtlebits 11 days ago
    Coming from a devops background, I kind of wish they didn't support arbitrary docker containers, as I found the experience fairly painful (going through one of their guides), and soured me from using Fly at all. Maybe it was the app's fault, but troubleshooting was not intuitive at all, especially on the networking side.
    • TillE 11 days ago
      Unless you're doing something that is explicitly supported with flyctl's automatic setup, I found writing a basic Dockerfile was far far easier than messing with buildpacks.
    • digianarchist 11 days ago
      You'd prefer if they only supported build-packs/nix-packs? I can see the appeal from a simplicity standpoint but supporting containers expands the usecase significantly.
  • site-packages1 11 days ago
    I had a good experience doing the test app with fly, I’d definitely consider it for something internet routable and non-complex like a bot or something. Very excited for when I can run more complicated workloads such as workloads started with Docker Compose, without having to implement all the Docker Compose functionality myself.
  • bergenty 11 days ago
    I’ve been seeing all over this site, definitely have to check them out going by their blog posts.
  • andy800 10 days ago
    I think would be a great platform for my apps but sadly they do not support the JVM. Sure, I could Dockerify my apps but at that point a simple VPS isn't much different.
  • mhh__ 11 days ago
    Fly made infrastructure really hard for me because their hypervisors cpuid emulation broke my program.

    Yes, fly employees, I will file a bug somewhere - or email me.

    • mrkurt 11 days ago
      That's interesting. I don't think we've run across that before, definitely send us more info!
  • itake 11 days ago
    Does fly recommend any good tools for measuring client latency (which is their big selling point)? I know they offer graphs for server latency.
    • mrkurt 11 days ago
      So far, people seem to get the most out of the `performance.timing` API and a metrics package. There are very expensive real user monitoring tools to do some of this, but that browser API makes it really easy to just wire up an app and collect whatever you care about directly.
      • itake 11 days ago
        Unfortunately, many of my users are in Asia. It is challenging to audit their user experience.
  • rickette 10 days ago
    Anyone experience running Java/JVM apps on Fly?
    • rurban 10 days ago
      Their free tier is limited to 256Mb, so'll need a bigger one.
  • ransom1538 11 days ago
    Ah. Full circle back to heroku.
  • jillesvangurp 11 days ago
    Docker containers are good deployment unit. At this point producing them and publishing them to a container registry is pretty straight forward. That's not something that requires devops people. Any reasonably senior backend developer ought to know how to do this or can learn how to do it in a few hours.

    The rest is just deploying and running containers. There are lots of ways to do that. I loved using Google Cloud Run a few years ago. Stupidly easy to get started with and flexible enough for many things. With some service discovery on top, it's perfect for a lot of stuff. Add some managed middleware & databases to the mix and you essentially have a close to zero ops CI/CD capable environment. No devops needed for this either. When I did this for the first time, I was up and running with our dockerized app in about 15 minutes. Most of that was just waiting for builds to finish.

    I'm CTO of a company currently and I've gotten sidetracked with enough lengthy and super expensive devops type stuff in past projects that I'm on purpose avoiding to go near certain things not because I can't do it but because I don't think these things are worth spending any time on for us right now. So, no terraform, no kubernetes, no microservices. I just don't have the time or patience for that stuff. We run a monolith. So, there's not a lot I actually need from my infrastructure. I need it to be fast, secure, and resilient and be able to run my monolith. But I don't need to have things like service discovery, complicated network setup (bog standard vpc is fine), and all the other stuff that devops people obsess about.

    We use a load-balancer, I clicked one together in the Google UI. It's fine. Ten minute job. Doesn't need terraform scripting. We have two of them. And we have a couple buckets and our monolith behind that. I could grab the gcloud command that recreates this thing and put it somewhere. But I have more urgent things to do.

    For deployment we use simple gcloud commands from github actions to update vms with new instance templates to tell them to run the latest container that our build produced. We started with cloud run but our monolith has a few worker threads that we don't want killed so we moved it to proper vms. Very easy to do in Google Cloud.

    Our deploy command does a rolling restart. We have health checks, logging, monitoring, alerting, etc. Could be better but it works. Initial provisioning of the environment was manual and we scripted together all the commands that are part of our deploy process for automation. We added a managed redis, database, and elasticsearch to this. None of that was particularly hard or worth automating to me. Yes, it's bit of a snowflake. But not that complicated and I documented it. So, we can do it again in a few hours if we ever need to.

    The dirty little secret of a lot of devops that it's a lot over over-engineered YAGNY stuff that is super labor intensive to setup and maintain and you end up using it a lot less often than people think.

    This is why freelance devops engineers are so in demand: this stuff just requires a lot of manual work! Companies need these people full time and usually more than just one. The devops alone can add up to hundreds of thousands of dollars/euros per year.

    It's a lot of manual work that probably should be automated. However, hiring a lot of people at great expense to automate things that are cheap and not that complicated is not always the best use of resources. I've seen companies that spend an order of magnitude more on devops salaries than on the actual hosting bills. If you think about it, that's kind of weird to be spending so much for so little gains. And most of these companies are not particularly big or experience enormous scaling issues.

    • strzibny 11 days ago
      I 100% agree with you and it's what I advise in the Scaling chapter of my Deployment from Scratch book.

      This is especially true when you realize you want a QA env, US/EU env, staging, etc. If it's one server (or server and DB), it's much much easier to create more environments.

  • Scarbutt 11 days ago
    Without PITR for postges I don’t think so.
  • newaccount2021 11 days ago
  • jononomo 11 days ago
  • melony 11 days ago
    There is nothing easy about them, cheaper yes (only for the lower instances), but not easier. Until they have first-class, evergreen, support for Heroku buildpacks, they are a subpar replacement.

    They are also missing a proper managed database with point-in-time backups through the web UI like those offered by most proper PaaS services.

    • mrkurt 11 days ago
      We aren't going to support Heroku buildpacks. We are working to make more frameworks easy to launch with minimal configurations. Which means – hiring people to work specifically on those frameworks + docs + builds.

      We did bake nixpacks into our CLI recently, they seem better for our particular environment than buildpacks. did a great job with these:

      We're working on managed databases, but we're not doing them like Heroku did. We just launched a preview of managed Redis with Upstash:

      This seems like the future of managed databases on a platform like ours. There are companies that build very good managed database services. We're getting to the size where these people will work with us. Getting well managed DBs onto the platform is basically what I'm spending all my time on these days.

      Incidentally, we're a lot cheaper than Heroku because we run our own infrastructure.

      • samwillis 11 days ago
        I’m sure you already have something in the pipeline for fully managed Postgres on Fly (I Really want something Heroku Postgres like, but on Fly). If not, have you reached out to Crunchy Data? They seem to have a solid product.

        If you happen to have a service beta testing anything I would be interested in joining it.

      • maxioatic 11 days ago
        Do you have plans for Rust (and specifically web frameworks like actix-web)? I see Rust on your main site under "Use the Tech You Love" but there's no dedicated docs.
        • mrkurt 11 days ago
          Yep! If all goes well, we'll have people working on Rust "last mile" UX sometime in 2023. We all use Rust, we just so happen to write Dockerfiles when we do.
          • maxioatic 10 days ago
            That's awesome! That's how I've used it so far and it works great, no complaints.

            ps. are those people already at fly? shameless plug - would love to work on things like that.

    • pyentropy 11 days ago
      > point-in-time backups through the web UI

      Is `fly postgres create --name restoredDb --snapshot-id backupId` that hard that's it's a deal breaker?

      > support for Heroku buildpacks

      I haven't tried it but there's some buildpack support:

      • melony 11 days ago
        It's not a managed database, if things break at 3am, you have to fix it yourself. The buildpacks are an afterthought using third-parties. A real evergreen solution will track the official Heroku solution (since doesn't even bother to document their buildpacks) to the nearest day. The whole point of the P in PaaS is that somebody else does the DevOps. If I have to do so much DevOps, then there's no point in adopting Fly. I am getting the impression that users here only care about their shiny blogposts and content marketing, versus the actual experience as a business user.
        • mrkurt 11 days ago
          I'm going to make a point here that might be contentious: you are asking for Heroku. Heroku is going away because it no longer works as a business.

          We are not Heroku. It is ok for you to not like what we're doing. We're building something different. We've never even _said_ we were a Heroku alternative, we just liked their UX for deploying apps and decided to roll with something similar.

          • victop 11 days ago
            > Heroku is going away because it no longer works as a business.

            I would rather say 'It no longer works as a business for Salesforce.'

          • ericpauley 11 days ago
            While I can't agree with melony on buildpacks (shipping OCI containers straight from our CI to Fly has been a game changer), I do agree on the importance of fully-managed databases.

            Our app has two components, a backend on Heroku and distributed frontend on Fly. The backend relies on a managed database, and I have not had to touch it in 6 years. Heroku does a great job providing confidence that the managed database will Just Work. The current Fly Postgres offering doesn't provide this confidence.

            • mrkurt 11 days ago
              I also agree about the importance of fully managed databases!

              We shipped "automated" Postgres because we couldn't get any fully managed DB providers to pay attention to us when we were small. I expect we'll have an option running on Fly infrastructure in the next six months.

              Our Redis is fully managed, so you can get an idea of how it might play out:

    • pier25 11 days ago
      It's not Fly's fault if Heroku decided to remove their free tier. They do have buildpacks for the most popular stacks. Eg: You can deploy a Node app really with just `fly launch`.

      As for the managed DBs, one of their founders was from Compose, so yeah they know how these things work. But AFAIK Fly doesn't have much interest in DBs, their focus is really in VMs.

    • Karupan 11 days ago
      Looks like you just want Heroku. And I’ve never seen the Fly team mention that they are aiming to be a Heroku replacement. So maybe it’s just not for you?