Argus: A Real-EKS Study of When Predicting Spot Interruptions Beats Simple Checkpointing

Excerpt

arXiv:2609.39067v1 Announce Type: new Abstract: Elastic Compute Cloud (EC2) Spot is 60% to 90% cheaper than On-Demand but can be reclaimed on just a 2-minute notice; for expensive multi-node training this loss can be severe, with one reclaim costing hours of synchronous progress. We build Argus, a Kubernetes operator, and ask empirically, on a CIFAR-10 testbed, when predicting interruptions beats simple checkpointing. Argus on real EKS survives a real Spot drain with a graceful SIGTERM checkpoin