The Zero-Touch Data Center: Lights-Out Operations at Scale

Abstract

Research shows that over 90% of the critical incidents that occur in hyperscale datacenters are related to same event type and with potentially same hardware a.k.a. host. In this paper, its described a Zero Touch Datacenter framework, Self Healing Architecture (SHArch) that takes away a majority chunk of work from the IT Operations team, responsible to maintain hyperscale datacenters. The framework orchestrates and heals these events and provides a consolidation process to further analyze noisy hosts for a permanent solution to these recurring events. SHArch has been extended to support IBM z/OS mainframe environments, enabling true lights-out operations across hybrid x86 and mainframe fleets. It achieves higher efficiency, brings end to end visibility in the entire operations process in real time. It is truly autonomous as it scales, remains highly available and remains flexible for adopting to continuously changing fault patterns.