Experiment: Try using b-linux-large workers for Taskcluster build tasks

[oskirby]

Description

We have a larger GCP worker available to us on Taskcluster. While we aren't frequently blocked by taskcluster jobs, it might be nice to get an idea of how much faster they could be. Especially if we have dreams of doing more testing in taskcluster.

Reference

None.

Checklist

• [ ] My code follows the style guidelines for this project
• [ ] I have not added any packages that contain high risk or unknown licenses (GPL, LGPL, MPL, etc. consult with DevOps if in question)
• [ ] I have performed a self review of my own code
• [ ] I have commented my code PARTICULARLY in hard to understand areas
• [ ] I have added thorough tests where needed

[oskirby]

Well, that is pretty cool. This resulted in faster build times by a factor of three! The question is, do we need it? IMO our development cycles right now are typically blocked by the functional tests as the weakest link so a faster Taskcluster build, while nice, doesn't really speed us up by as much as we would like.

[strseb]

Well, that is pretty cool. This resulted in faster build times by a factor of three! The question is, do we need it? IMO our development cycles right now are typically blocked by the functional tests as the weakest link so a faster Taskcluster build, while nice, doesn't really speed us up by as much as we would like.

wooow. Well then we might need to check the economics, the android runners already require us to book worker hours. If the this worker is not more expensive by the factor of 3 this might make sense to use anyway.

The big one seems to be a n2-custom-32 core machine, so we'd need to ask releng how much a job costs there 😅