Nonparametric multivariate anomaly analysis in support of hpc resilience

Large-scale computing systems provide great po-tential for scientific exploration. However, the complex-ity that accompanies these enormous machines raises challeges for both, users and operators. The effec-tive use of such systems is often hampered by fail-ures encountered when running applications on systems containing tens-of-thousands of nodes and hundreds-of-thousands of compute cores capable of yielding petaflops of performance. In systems of this size fail-ure detection is complicated and root-cause diagnosis difficult. This paper describes our recent work in the identification of anomalies in monitoring data and sys-tem logs to provide further insights into machine status, runtime behavior, failure modes and failure root causes. It discusses the details of an initial prototype that gath-ers the data and uses statistical techniques for analysis. 1