Heroku Autoscaling done rightwith queue time instead of response time
Explore for yourself in our sandbox app →
“Judoscale’s autoscaling approach based on request queue time is far superior to Heroku’s outdated approach based on response time. ”
Queue time over response time
Heroku’s response time autoscaling is unreliable. Judoscale uses queue time to autoscale your dynos, which is a much more accurate measure of capacity.Learn more about request queue time →
Autoscaling for standard dynos and more
Heroku only offers autoscaling for performance dynos. Judoscale works with all Heroku dyno types, including standard, performance, private, and shield dynos.
Worker dynos included
Autoscaling is as important for worker dynos as web dynos. Judoscale monitors your job queues (such as Sidekiq or Celery), scaling them before they get backed up. You can even scale down to zero dynos for workers that only run occasionally.
3x faster than Heroku's autoscaler
A capacity issue needs to trigger autoscaling as quickly as possible, and Judoscale is the fastest autoscaler available. Our autoscaling algorithm runs every 10 seconds, ensuring your app scales up before users notice an issue.
Take control of your Heroku bill
Most Heroku apps use way more dynos than necessary, earning Heroku a reputation for being too expensive. Judoscale optimizes your dyno usage, often reducing your Heroku bill by 50% or more.
Customize your autoscale behavior
Every app is different, and so Judoscale gives you complete control over how your app autoscales. You can scale by multiple dynos at a time and tweak the frequency of scaling. Each process is configured independently with a few simple sliders.Explore the configs in our sandbox app →
- Trusted by900+engineering teams
- Over2.5 millionautoscales per month
- Since2017we are here to stay
Got questions?
Check out our docs for a whole lot more. If you still can’t find what you’re looking for, send us an email!
What happened to Rails Autoscale?
We rebranded as Judoscale in 2022 to be more inclusive to other frameworks. The Rails Autoscale add-on exists for backward-compatibility, but it’s identical to the Judoscale add-on.
What languages and frameworks does Judoscale support?
We support many web frameworks and job/task queues for Ruby, Python, and NodeJS. See the full list in our docs.
Does Judoscale support Private Spaces dynos?
Yes! We support both Private and Shield dynos. Note that request and job metrics are sent to Judoscale over the public internet.
What data can Judoscale access in my app?
Our adapter libraries only collect queue-related metrics for requests and job queues along with basic process metadata. No actual request or job data is ever collected.
What’s included in the free plan?
Our White Belt plan is completely free, and you can use it forever. It’s limited to 20 autoscales per month, perfect for staging and demo apps. Learn more on our pricing page.
Can we have a call to see if Judoscale makes sense for our Heroku app?
Sure, let’s talk! Use this link to book a call with us.
Heroku’s response time autoscaling would always lead to response times spiking for a few minutes when it downscaled too much at once. Judoscale fixed that issue for us.
We have very large fluctuations in the load of our application, and we wanted to save money. Judoscales proved itself well, it reacts fast to the traffic spikes and Heroku dynos spin up quickly.
I feel like it’s been a bargain for the headaches it saves me. It lets us not have to dive into Heroku each time there’s a burst of activity.
Judoscale has been useful in that it lets us keep the number of dynos for processing relatively low with the possibility to surge when needed.
Our servers now happily scale anywhere from 2 to 15 dynos, saving us thousands a month.
Judoscale is tailor-made to align with the teachings in Nate Berkopec’s Complete Guide to Rails Performance. I signed up initially to scale our worker dynos, which worked exceptionally well. Then I took the plunge to auto-scale the web dynos, and we’ve never switched it off.
Start autoscaling for free
Setup takes less than 5 minutes