Documentation/How do I checkpoint my job
From systems
NOTE: This page is out of date.
Goal of checkpointing implementation in your applications
Some lengthy calculations can run for months. If the calculation node dies due to a
hardware failure or power loss, you will lose all the calculation time and have to
restart from the beginning. Having a checkpoint created, you can at least restart
from this point of calculation.
Go to this link to learn how to use checkpointing.