See new real world captures of datacenter applications at the file system and block IO level. Presentation and analysis of GPS Nav Portal and other workloads. See in situ performance of native datacenter servers. Compare in situ performance to lab testing of SSD and HDD servers using the actual real world workload captures.

Data stored in storage are often corrupted silently without any explicit error. To cope with the silent data corruption, the file system level detection is effective; Btrfs and ZFS have a mechanism to detect it by adding checksum in each block. However, data replication is required to correct the corrupted data, which waste storage capacity in local file system. Gfarm file system is an open-source distributed file system, which federates storages even among several institutions in wide area. It supports file replicas to improve access performance and also fault tolerance.

One of the most significant challenges for software defined storage systems is to determine an appropriate tuning configuration considering diverse factors including: hardware, software stacks, and workload characteristics. Currently, the configuration of complex storage systems relies on the experience and skills of human experts. For storage companies, this creates bottlenecks on customer service time, since the number of experts is limited. Moreover, the tuning expertise can be diminished by continuous personal rotation and the lack of experience sharing.

New IO Capture and analysis tools - such as those available for free at www.TestMyWorkload.com (TMW) - now enable users to capture, analyze and test Datacenter Real World Storage Workloads

A major component of request latency is the time back-end servers spend reading and writing data. Object storage systems commonly use JBOD servers together with a general purpose file system to store object data in files. Yet there are far more efficient ways of implementing the back-end storage. In this presentation, we explore the lessons we learned, problems we encountered, and resulting performance gains in the process of creating a new back-end storage format.

The requirement for ever denser memories is answered by the introduction of new NAND memory technologies. New multi-layer three dimensional stacking (BiCS4) as well as four bits per cell (QLC) technologies are introduced by the NAND memory FAB companies. The introduction of these technologies brings into bear reliability challenges which are answered with new and unique methods. Any memory controller must contend with the variability between the layers in 3D manufacturing technologies as well as the high error rates and stress condition sensitivity introduced by QLC.

Applying micro service patterns to storage giving each workload its own Container Attached Storage (CAS) system. This puts the DevOps persona within full control of the storage requirements and brings data agility to k8s persistent workloads. We will go over the concept and the implementation of CAS, as well as its orchestration. Learning Objectives: 1. Go over the modern day apps and their storage needs; under the notion of applications have changed someone forgot to tell storage 2.

We are on the precipice of a new paradigm in the data center. Booming growth in both compute and storage have effectively changed how best to tie the traditional SW and HW layers together for scale, as well as which new SW and HW layers to use. The industry has answered the call with strategic changes in everything from storage device functionality, form factors and capacity points, to new compute platforms and the HW used to perform that compute. New offerings of open source HW and SW, tied together with new more functional and powerful networks help finish this new puzzle.

Two key trends are driving the need for computational SSDs. The first trend is the data-heavy nature of modern workloads. The best machine learning algorithms will perform poorly in the absence of large amounts of high quality data. Similarly, analytics weather video, log or database all require a massive amount of data. The second key trend is the diversity of compute workloads like machine learning, big data analytics, and streaming video in the modern datacenter, none of which are a great fit for the CPU instruction set.

The Storage Networking Industry Association’s (SNIA) Solid State Storage Initiative (SSSI) has released the first version of its Solid State Storage Performance Test Specification (SSS PTS)i . The SSS PTS, combined with a good reference test system, is an important tool for system integrators to use to find the right solid state drive (SSD) for their intended application. This white paper illustrates how the SSS PTS can effectively be used to compare a number of SSDs. The paper describes test results from 17 SSDs currently on the market.