SNIA Podcasts

Storage Developer Conference Podcast Series

Every week the Storage Developer Conference (SDC) podcast series presents important technical topics to the Storage Developer community. Each episode is hand-selected by the SNIA Technical Council from the presentations at SDC.

This week’s highlighted Podcast: #147: Platform Performance Analysis for I/O-intensive Applications by Ilia Kurakin, Intel, and Perry Taylor, Intel  (MP3, Video, PDF of Slides)

High-performance storage applications running on Intel® Xeon® processors actively utilize I/O capabilities and I/O accelerating features of the platform by interfacing with NVMe devices. Such I/O-intensive applications may suffer from performance issues, which in a big picture can be categorized into three domains: I/O device bound – performance is limited by device capabilities core bound – performance is limited by algorithmic or microarchitectural code issues uncore bound – performance is limited by non-optimal interactions between devices and CPU. This talk focuses on the latter case. In Intel architectures, the term “core” covers execution units and private caches, and all the rest of the processor is referred as “uncore”, which includes on-die interconnect, shared cache, cross-socket links, integrated memory and I/O controllers, etc. Activities happening on the IO path in uncore cannot be monitored with traditional core-centric analyses, but there are pitfalls that require an uncore-centric view. Intel servers provide such a view by incorporating thousands of uncore performance monitoring events that can be collected in performance monitoring units (PMUs) associated with uncore IP blocks. However, using raw counters for performance analysis requires deep knowledge of hardware and appears incredibly challenging. In this talk, we will discuss platform-level activities induced by I/O traffic on Intel® Xeon® Scalable processors and summarize practices for best performance of storage applications. We will overview telemetry points staying on the IO traffic path and eventually present developing uncore-specific performance analysis methodology, that reveals platform-level inefficiencies, including poor utilization of Intel® Data Direct I/O Technology (Intel® DDIO).

Learning Objectives: Uncore-centric performance analysis methodology for I/O-intensive applications running on Intel server architectures, HW operations induced by PCIe traffic, and HW-level observability for them; Practices to gain best IO performance on Intel server architectures.

The Podcast can be downloaded on iTunes. Download a copy of the original presentation slides as well.



Note: SNIA Podcasts are provided as information only. Articles and guests that appear on any SNIA programs express the opinions, experience, technical positions, and research of those authors and guests and not necessarily those of the SNIA.

SNIA Podcasts Copyright 2007 -  2020. All rights reserved. No part of any SNIA Podcast may be reproduced, stored in a retrieval system, or transcribed, in any form or by any means, photocopying, electronic, mechanical recording or otherwise for personal or professional financial gain, without prior permission in writing of the publisher, The Storage Networking Industry Association, 4360 ArrowsWest Drive, Colorado Springs, CO 80907 USA