Rebasing Private Spaces with Kubernetes

[friism]

We're looking to expand Heroku's opinionated developer experience by shifting the Heroku strategy to be based on Kubernetes. By doing so, we will be able to provide improvements to the core Private Spaces product such as:

• Faster dyno boot times
• Removing Daily Dyno Restarts
• More flexible (and smaller minimum) dyno sizes
• Performance during code deploys due to how dynos are currently cycled in Private Spaces
• Dyno scale limits

In addition, by building on Kubernetes and therefore bringing Heroku closer to open standards and cloud-native technology, we can unlock new platform enhancements currently captured on the public roadmap today like:

Native OpenTelemetry on Heroku

and

Cloud Native Buildpack Support

Required Terms

• I agree to follow this project's Code of Conduct
• I have read and accept the Salesforce Program Agreement

[JohanBengtsson]

Hi there, has there been actual research done on this ticket? Is there anything you can share on your progress?

[elimchaysengSF]

I don't have specific details or timelines I can share as of right now, but these specific improvements are something actively being worked on now that are being wrapped into a larger Heroku initiative. There should be some more public information coming within the next few weeks and being discussed at TDX this week that addresses these improvement and the future vision of the Private Spaces product.

[plehoux]

@elimchaysengSF

For us, the shortcomings of Heroku private space runtime versus the common runtime mean that since we moved to private space, we are using 25% more dynos. This could be addressed, of course, but with increased complexity.

Additionally, our Heroku/Salesforce rep is pushing us to renew our annual contract early or risk facing a 9% increase upon renewal.

This roadmap entry is interesting because we were strongly considering moving to AWS bare-bone.

At TDX this week that addresses

Has anything been announced yet? I could not find any new information.