Comparison of prices to Heroku


Am I correct in analysing the pricing between Heroku and Fly.io to say that a Standard 1 dyno, 512MB Ram at $25 a month is equivalent to $3.19/mo on Fly.io?

What is the underlying tech that fly.io is hosted on? (Heroku was AWS I believe)

Is there any particular downside to being on a share box over a dedicated CPU? So far on Heroku I haven’t really seen any issues.

TIA!

From what’s publicly available: OVH, Equinix, NetActuate, CacheNet (ref). Fly uses AWS (S3, RDS etc), but for control-plane, iiuc.

Kurt says, pound-for-pound dedicated-cpu on Fly is faster: Heroku equivalents for PL Dyno - #2 by kurt

With shared-cpu, of course, there’s only a limited timeslice available to a VM, which is ~1/15th for Fly’s lowest tier, share-cpu-1x: Heroku Free Tier Migration - #6 by ignoramous

1 Like

I wrote a small comparison at Migrate from Heroku · Fly Docs for Rails apps that can give you a rough idea of how costs are different. Hope that helps!

1 Like

It does not really.

Am I correct in analysing the pricing between Heroku and Fly.io to say that a Standard 1 dyno, 512MB Ram at $25 a month is equivalent to $3.19/mo on Fly.io?

Or do I need to be on a dedicated CPU because the shared ones on fly are not comparable to heroku shared ones?

The prices difference is significant so I am dubious!

I like the comparison. One suggestion I have is to rename the “Resource” header to either “Heroku Resource” or Fly.io Resource” depending on the table. I would assume most folks will skim the tables and it’s not readily apparent which table is for which service unless you read the paragraphs beforehand.

2 Likes

@RobertD yes, a shared-cpu-1x with 512MB RAM is equivalent to Heroku’s standard dyno.

Heroku is extremely expensive. This is mostly because they run on top of AWS. We run on our own hardware, and can bill roughly what AWS does for VMs as a result.

If you compare our pricing to, like, t2 instances on AWS you’ll see that we’re a little more expensive. We hold up reasonably well to DigitalOcean, too.

2 Likes

Great point—change made and deployed