We are in the process of migrating from Heroku.
We are currently on the Private Spaces plan, so our Dynos are of “Private” type:
So looking at those private-(s/m/l), it’s easy to find the equivalent regarding memory-wise.
But the problem is, how would you compare CPU?
So the question is, what is the equivalent of Heroku’s private-(s/m/l) in Render, regarding CPU power?
We replied through another channel but repeating it here for completeness.
It’s really hard to compare services specifically when it comes to compute. RAM is easier as that’s a finite amount of resource - but compute is tough. Platforms don’t provide great explanations of what the figures actually mean - from your Heroku figures, what does 1x CPU share actually mean and the compute column note states:
Overall, performance varies significantly based on app implementation. These figures show expected performance based on 99 percent of historical system loads. The performance of apps running on
standard dyno types will vary based on available system resources.
So the short answer is that we can’t compare ourselves to Heroku in terms of CPU - it’s going to be a hand-wavy fluffy “It depends” answer.
Depending on the resource requirements of your own application you should use the most suitable Render plan - that’s going to come down to benchmarking your service to see how it performs and made decisions accordingly. Our 1GB RAM plan may have less compute resource available so where you were using a 1GB RAM plan on Heroku you might need to step up to a higher plan on us to have more compute etc.
This topic was automatically closed 30 days after the last reply. New replies are no longer allowed.