Assuming Failure Independence: Are We Right to be Wrong?

This paper revisits the failure
¹
temporal independence hypothesis which is omnipresent in the analysis of resilience methods for HPC. We explain why a previous approach is incorrect, and we propose a new method to detect failure cascades, i.e., series of non-independent consecutive failures. We use this new method to assess whether public archive failure logs contain failure cascades. Then we design and compare several cascade-aware checkpointing algorithms to quantify the maximum gain that could be obtained, and we report extensive simulation results with archive and synthetic failure logs. Altogether, there are a few logs that contain cascades, but we show that the gain that can be achieved from this knowledge is not significant. The conclusion is that we can wrongly, but safely, assume failure independence!