Timeouts: Add support for more than pipeline

Just like for CI-Tron jobs, there is a lot of value to having fine-grained timeouts:

• Quick reaction time to hangs (watchdog-style)
• Smaller worst-case execution time when retries are necessary
• ...

To this end, we should add timeouts for every part of the pipeline, whether repeated (pulling/running containers) or entire sections (pipeline setup/teardown, volume mirroring, ...)