A Recovery Conscious Framework for Fault Resilient Storage Systems

In this paper we present a recovery-conscious framework for improving the fault resiliency and recovery efficiency of highly concurrent embedded storage software systems. Our framework consists of a three-tier architecture and a suite of recovery conscious techniques. In the top tier, we promote fine-grained recovery at the task level by introducing recovery groups to model recovery dependencies between tasks. At the middle tier we develop highly effective mappings of dependent tasks to processor resources through careful tuning of recovery efficiency sensitive parameters. At the bottom tier, we advocate the use of recovery-conscious scheduling by careful serialization of dependent tasks, which provides high recovery efficiency without sacrificing system performance. We develop a formal model to guide the understanding and the development of techniques for effectively mapping fine-grained tasks to system resources, aiming at reducing the ripple effect of software failures while sustaining high performance even during system recovery. Our techniques have been implemented on a real industry-standard storage system. Experimental results show that our techniques are effective, non-intrusive and can significantly boost system resilience while delivering high performance