For now there's only one hard timeout for wait-for-boot: if 10 nodes are required in an experiment and 9 of them boot in 30s, the actual behaviour will keep waiting until max-wait (2 minutes) even if nothing happens.
It would be better to have a soft and hard timeout:
- soft timeout: if nothing happens within
t, the program returns
- hard timeout: skip any node that didn't boot after hard timeout (even if there is still some activity)
For now there's only one hard timeout for wait-for-boot: if 10 nodes are required in an experiment and 9 of them boot in 30s, the actual behaviour will keep waiting until max-wait (2 minutes) even if nothing happens.
It would be better to have a soft and hard timeout:
t, the program returns