The scheduled docker build+deploy action has been failing the last 3 cycles.
This first looked like a problem with the docker setup itself, because it was complaining of running out of space inside the container, but it turns out that it is actually the GitHub runner VM that is running out of space. It seems that the sum of all pulled images + installs is more than the runners have available.
I can see two options forward:
move the action to an AWS VM that has more local disk space
split up the action into two separate workflows so that the runner VM can recover disk space in between. That would hopefully be enough.
The scheduled docker build+deploy action has been failing the last 3 cycles.
This first looked like a problem with the docker setup itself, because it was complaining of running out of space inside the container, but it turns out that it is actually the GitHub runner VM that is running out of space. It seems that the sum of all pulled images + installs is more than the runners have available.
I can see two options forward: