Combining The Linux Infrastructures


We’re combining the linux infrastructures into a single one. Read away on our blog for all details and feel free to leave a comment. :slight_smile:


Will the new infrastructure still be centered around the ubuntu distributions? I noticed that the default one is currently still trusty, which reaches eol next year, while the next LTS one is still marked as experimental. So I wonder if it makes sense to prepare for Xenial, or if it will go into another direction.


There are no plans as far as I’m aware to step away from Ubuntu distributions, so I’d continue planning around that. We’ll be sure to let you know if that changes! :slight_smile:


Given that the container-based infrastructure boots faster, how will this impact the end-to-end speed of our builds?


We’re working on improving this in a number of ways. Among them, we’re adding pre-booted VMs, and making further resource improvements. We’ll share updates as soon as we can, but we are aiming to bring boot time as low as possible. :slight_smile:


In your article: it states:
" Over the next few weeks, we encourage everyone to remove any sudo: false configurations from your .travis.yml . Soon we will run all projects on the virtual-machine-based infrastructure, the sudo keyword will be fully deprecated."

Most, if not all, of our projects use sudo: required. Should we leave this as is, or remove sudo completely?


You can leave it as is, if you’d like. If you’re updating the configs for other reasons feel free to remove it, but there’s no need to go out of your way. :slightly_smiling_face: