LVC21F-107 Unleashing the performance of multi-actuator drives

Session Abstract

Level: Intermediate  Reaching higher IOPS becomes increasingly important as drive capacities grow. Multi-actuator drives are an effective response to this need. They appear as a single device to the I/O subsystem. Yet they address commands to different actuators internally, as a function of LBAs. This poses the following important challenge: information on the destination actuator of each command must be used cleverly by the I/O subsystem. Otherwise the system has little or no control over the load balance among actuators; some actuators may be underutilized or remain totally idle. I/O schedulers are the best candidates for tackling this problem, as their role is exactly to decide the order in which to dispatch commands. In consultation with Seagate Technology, we have enriched the BFQ I/O scheduler with extra logics for multiple actuators. Exploiting knowledge of the destination actuators of commands, this extended version of BFQ provides dramatic performance improvements, over a wide range of workloads. At the same time, it preserves the original bandwidth and latency guarantees of BFQ. As a more general contribution, the concepts and strategies used in BFQ show effective ways to take advantage of the IOPS gains of multi-actuator drives.

Session Speakers

Paolo Valente

Linaro, Collaborator, Assistant professor

Paolo Valente is an Assistant Professor of Computer Science at the University of Modena and Reggio Emilia, Italy, and a collaborator of the Linaro engineering organization. Paolo's main activities focus on scheduling algorithms for storage devices, transmission links and CPUs. In this respect, Paolo is the author of the last version of the BFQ I/O scheduler. BFQ entered the Linux kernel from 4.12, providing unprecedented low-latency and fairness guarantees. As for transmission links, Paolo is one of the authors of the QFQ packet scheduler, which has been in the Linux kernel until 3.7, after that it has been replaced by QFQ+, a faster variant defined and implemented by Paolo himself. Finally, Paolo has also defined and implemented other algorithms, part of which are now in FreeBSD, and has provided new theoretic results on multiprocessor scheduling.

Tim Walker

Principle Engineer (Seagate)

I work with our key customers integrating our new technology into their storage systems.

Level: Intermediate 

Reaching higher IOPS becomes increasingly important as drive capacities grow. Multi-actuator drives are an effective response to this need. They appear as a single device to the I/O subsystem. Yet they address commands to different actuators internally, as a function of LBAs. This poses the following important challenge: information on the destination actuator of each command must be used cleverly by the I/O subsystem. Otherwise the system has little or no control over the load balance among actuators; some actuators may be underutilized or remain totally idle. I/O schedulers are the best candidates for tackling this problem, as their role is exactly to decide the order in which to dispatch commands. In consultation with Seagate Technology, we have enriched the BFQ I/O scheduler with extra logics for multiple actuators. Exploiting knowledge of the destination actuators of commands, this extended version of BFQ provides dramatic performance improvements, over a wide range of workloads. At the same time, it preserves the original bandwidth and latency guarantees of BFQ. As a more general contribution, the concepts and strategies used in BFQ show effective ways to take advantage of the IOPS gains of multi-actuator drives.

comments powered by Disqus

Recent Posts

Other Posts

Sign up. Receive Updates. Stay informed.

Sign up to our mailing list to receive updates on the latest Linaro Connect news!