mrkurt

53 ms·

601.
by mrkurt 6y ago
Yep! Our internal service discovery lets you connect to instances in the same region. If you're running in "SJC" you could use "sjc.my-postgres-cluster.internal", for example.
602.
by mrkurt 6y ago
It's definitely not naivete or stupidity, but projects from large companies inherit all the complexity of the organization that created them. This is sometimes because they're very broadly scoped, and sometimes because large group
603.
by mrkurt 6y ago
$0.02/GB is higher than we want to charge, but we have to pay the same rates you do at most providers. It's a little unfair to call it "highway robbery", though, we have to price things so we can stay alive. We're p
604.
by mrkurt 6y ago
Yes. The advantage of Firecracker is that it's lightweight and boots incredibly fast. It's a pretty good way to run containers. But the virtualization primitives are almost the same.
605.
by mrkurt 6y ago
It's somewhere between Heroku and DigitalOcean. We have a CLI for managing and deploying apps that's very similar to Heroku. But apps also get private networking, disks, load balancing for TCP/UDP. We do have people deploy si
606.
by mrkurt 6y ago
I'm glad you noticed. It sure makes our lives simpler.
607.
by mrkurt 6y ago
Yes. Our beta postgres service lets you add read only replicas in any region you want. "Deploy any full stack app globally" has been a big goal for us over the last year. We're getting close.
608.
by mrkurt 6y ago
The simple answer is: because containerd + devicemapper makes it easy and we already use LVM. We use LVM so we can expose raw block devices to peoples' applications.
609.
by mrkurt 6y ago
I'm one of those weird people who thinks the world would be better if acquisitions didn't happen. Some of my best work is gone from the internet because a large company acquired it and let it rot.
610.
by mrkurt 6y ago
You're getting downvoted but I don't think that's fair. Some people want social proof, we're not going to get them as customers. There's a reason "no one ever got fired for buying AWS". I will say that mos
611.
by mrkurt 6y ago
This is how it works for us. When we hire a marketing group, we'll work through the process, do case studies, etc. But it's more fun to write technical content.
612.
by mrkurt 6y ago
I'm surprised how well liveview copes with a flood of events. I've seen the latency indicator at the bottom spike at times, but it mostly just hums along.
613.
by mrkurt 6y ago
I have been surprised how little good tooling there is for Elixir, and even full stack in general. Clustering is supported, you can set libcluster to use our internal DNS to discover nodes: https://github.com/fly-apps/p
614.
by mrkurt 6y ago
Right? It's ridiculous.
615.
by mrkurt 6y ago
Not yet, we want to make sure `fly launch` does a good job with Elixir apps first. We're close.
616.
by mrkurt 6y ago
I am lucky enough to have read the drafts. These articles are a huge grind. Most of the skill seems to be putting the time in and having enough experience to go as deep as necessary.
617.
by mrkurt 6y ago
Well hello addison! Fancy meeting you here. It is very cool what's available to mashup these days. Kind of a lot to keep up with sometimes.
618.
by mrkurt 6y ago
I haven't tried the Java driver, but it's worth a shot. They also have "raw_exec" which might be helpful.
619.
by mrkurt 6y ago
That's exactly why we used nomad. It's focused and easy to understand, and we can build on top of it. We've even patched it on occasion. We manage LVM ourselves, yes. We don't do any live migration. Our disk offering is
620.
by mrkurt 6y ago
Specifically, containerd with the device mapper plugin!
621.
by mrkurt 6y ago
Check out the Nomad exec driver. It can orchestrate processes outside of containers, and even knows how to pull builds from a URL. Nomad is great. https://www.nomadproject.io/docs/drivers/exec
622.
by mrkurt 6y ago
When we started with Firecracker, there weren't really any projects built on top of it. Even if they had been, though, most things that touch kubernetes are incredibly complex. We have to be very careful how we manage complexity. For e
623.
by mrkurt 6y ago
You might get a kick out of this: https://liveview-counter.fly.dev/ It's a multi region Phoenix app using libcluster to power LiveView. We have some fun ideas for Elixir, we just need to speed up builds.
624.
by mrkurt 6y ago
We're happy Sidekiq Pro customers too.
625.
by mrkurt 6y ago
I've also been rejected by YC several times! I think not doing YC is a really smart choice for some companies. That said, I think the value of yc is twofold. (1) it's the best possible way to get to a first round of funding and (2
626.
by mrkurt 6y ago
You need private networking or the Erlang nodes can't talk to each other.
627.
by mrkurt 6y ago
Mostly just random projects with my brother. Most of them go nowhere but it's a clean way to kick things off that involve money. Why is Wyoming the default, out of curiosity?
628.
by mrkurt 6y ago
I love this because I need, like, two LLCs per year for random, probably short lived, things.
629.
by mrkurt 6y ago
Love your app.
630.
by mrkurt 6y ago
Improv grew up especially gross, there's a lot of baggage to get rid of. It wasn't gross in a "this isn't PC" kind of way, it was gross in a "the people who were largely responsible for the rise of improve were

More ›