The life cycle of any physical device is a series of transitions from an initial healthy state to ultimate failure. Since Markov chains are a general method for modeling state transitions, they can effectively model the transitions as well as life expectancy. As a specific example, HDD life cycle can be analyzed by developing a Markov model from various disk indicators (e.g. medium errors, RAS, usage). An illustration is given wherein the evolution of medium errors in HDDs is modeled by an absorbing Markov chain.

Persistent Memory exhibits several interesting characteristics potentially including persistence, capacity, byte addressability, and others. This session explores the various characteristics usually mentioned in a discussion of persistent memory, and relates them to how applications might leverage those characteristics.

Persistent memory will significantly improve storage performance. But these benefits can be harder to realize in distributed storage systems such as Ceph or Gluster. In such architectures, several factors mitigate gains from faster storage. They include costly network overhead inherent for many operations, and deep software layers supporting necessary services. It is also likely that the high costs of persistent memory will limit deployments in the near term. This talk shall explore steps to tackle those problems.

Persistent Memory (PM)-based file systems are key to bringing the low-latency value of PM hardware to applications and businesses. But developing a portable POSIX-compliant file system using the traditional Kernel-to-user space FUSE bridge results in high-latency access, and thus defeats the purpose. To bridge this gap NetApp has developed the open source ZUFS bridge. ZUFS was designed with PM latencies and byte addressability in mind.

As storage devices get faster, data management tasks rob the host of CPU cycles and DDR bandwidth. In this session, we examine a new interface to storage devices that can leverage existing and new CPU and DRAM resources to take over data management tasks like availability, recovery, and migrations. This new interface provides a roadmap for device-to-device interactions and more powerful storage devices capable of providing in-store compute services that can dramatically improve performance.

Free-cooling lowers datacenter costs significantly, but may also expose servers to higher and more variable temperatures and relative humidities. It is currently unclear whether these environmental conditions have a significant impact on hardware component reliability. Thus, in this paper, we use data from nine hyperscale datacenters to study the impact of environmental conditions on the reliability of server hardware, with a particular focus on disk drives and free cooling. Based on this study, we derive and validate a new model of disk lifetime as a function of environmental conditions.

Enterprise storage RAS (Reliability, Availability & Serviceability) solutions demands a highly-predictable and resilient storage strategy that enables 99.999% (5-nines) uptime.The right RAS storage strategy consists of the most appropriate hardware and software building blocks suited to the use case. The most well suited enterprise storage strategies must include a bottom-up RAS strategy, to build robust software on top of fundamental hardware extensions that allows for seemless fault recovery.

The tightly coupled architecture used by the vast majority of enterprises today is archaic and a new approach is needed to manage the explosion of data in a world of shrinking IT budgets. Enter: hyper-scale IT architecture.

Sudhakar Mungamoori will explain why a modern, software-driven and loosely coupled architecture is required for hyper-scale IT. He will highlight how this innovative architecture approach mitigates complexity, improves agility and reliability through on demand IT resources and reduces costs by as much as 10X.

Modern Software Defined Storage systems use heterogeneous storage media including disks, SSDs, and (soon) persistent memory. It’s important for these systems to identify, segregate and optimally utilize such storage media to gain advantages in performance, capacity and cost. Gluster's Data Tiering feature aims to help tackle these problems by segregating fast and slow storage into separate tiers, and providing intelligent movement of data across tiers according to dynamically observed current usage patterns.

Server resource allocation and traffic management is a large area of research and business concern in order to ensure proper functionality and maintenance procedures. As a result, good server reliability models that can incorporate workload and traffic stress are necessary. Here, we discuss server reliability models starting from the first dynamic proposal by Cha and Lee in 2011, then move to work by Traylor and Korzeniowski that better seeks to reflect scenarios found in actual servers.