Why are we paying per concurrent build on infrastructure **we** own and control?

[Noah Bailey]

I totally get paying per-minute for cloud compute - that's how it's always been, and nobody expects that to change. 

But charging us for using our own hardware is not a welcome change for anybody. We specifically used self hosted runners so that we can run long and intensive builds that would be too expensive to run in the cloud. And, like many orgs, we've optimized them by breaking up build and test steps into decoupled pieces that can be run in parallel. 

This feels, at least to me, like a senseless cash grab. There's no technical limitation on BB/Atlassian's end that prevents us from using our own hardware for concurrent builds. We're not even getting any improvement out of this, just paying more to do something we've already done for years. I could understand paying for advanced features, or even for the number of runners, but this is such a strange and arbitrary license limitation. It doesn't cost Atlassian anything for us to run our own builds!

https://www.atlassian.com/blog/bitbucket/announcing-v5-self-hosted-runners

https://support.atlassian.com/bitbucket-cloud/docs/configure-runner-concurrency-and-inspect-step-queue/

[Jonah Jacobsen]

We are in the same situation. Incredibly off-putting. This is the kind of thing that would prompt our entire org to migrate away from bitbucket

[Noah Bailey]

Yeah, similar situation here. We've resisted moving to github/gitlab because our team really likes the simpler interface and straightforward pipelines config, but this might push us to another solution if it will literally double or triple our costs overnight. The bean-counters surely won't be pleased with this...