Brendan Gregg

JULY 4, 2021

This was a chance to talk about other things I've been working on, such as the present and future of hardware performance. I also wrote about these topics in detail for my recent [Systems Performance 2nd Edition] book. Note that my predictions in this talk may be wrong, but they should be thought provoking. Ford, et al., “TCP

Performance 140

Performance Latency Hardware Storage 140

The Netflix TechBlog

JUNE 4, 2019

Because microprocessors are so fast, computer architecture design has evolved towards adding various levels of caching between compute units and the main memory, in order to hide the latency of bringing the bits to the brains. This avoids thrashing caches too much for B and evens out the pressure on the L3 caches of the machine.

Cache 251

Cache Latency Airlines Logistics 251

Sutter's Mill

FEBRUARY 13, 2017

Tue-Thu Apr 25-27: High-Performance and Low-Latency C++ (Stockholm). On April 25-27, I’ll be in Stockholm (Kista) giving a three-day seminar on “High-Performance and Low-Latency C++.”

Latency 51

Latency C++ Hardware Performance 51

ACM Sigarch

MAY 31, 2023

Introduction Memory systems are evolving into heterogeneous and composable architectures. Heterogeneous and Composable Memory (HCM) offers a feasible solution for terabyte- or petabyte-scale systems, addressing the performance and efficiency demands of emerging big-data applications. The recently announced CXL3.0

Latency 52

Latency Hardware Cache Architecture 52

Percona

NOVEMBER 2, 2023

MongoDB is a dynamic database system continually evolving to deliver optimized performance, robust security, and limitless scalability. Improved performance : MongoDB continually fine-tunes its database engine, resulting in faster query execution and reduced latency. Ready to supercharge your MongoDB experience?

Best Practices 86

Best Practices Hardware Tuning Scalability 86

The Netflix TechBlog

AUGUST 30, 2021

Complementing the hardware is the software on the RAE and in the cloud, and bridging the software on both ends is a bi-directional control plane. System Setup Architecture The following diagram summarizes the architecture description: Figure 1: Event-sourcing architecture of the Device Management Platform.

Latency 213

Latency Traffic Transportation Hardware 213

Brendan Gregg

FEBRUARY 28, 2023

This talk originated from my updates to [Systems Performance 2nd Edition], and this was the first time I've given this talk in person! CXL in a way allows a custom memory controller to be added to a system, to increase memory capacity, bandwidth, and overall performance. Ford, et al., “TCP

Performance 110

Performance Latency Cache Virtualization 110

The Morning Paper

JUNE 13, 2019

HotOS’19 is presenting me with something of a problem as there are so many interesting looking papers in the proceedings this year it’s going to be hard to cover them all! Perhaps inspired by serverless in spirit and in terminology, the path forward proposed in this paper is towards a managed and model-less inference serving system.

Hardware 58

Hardware Latency Serverless Energy 58

Dotcom-Montior

MAY 12, 2020

Websites are now more than just the storage and retrieval of information to present content to users. They now allow users to interact more with the company in the form of online forms, shopping carts, Content Management Systems (CMS), online courses, etc. Network latency. Hardware resources. Network Latency.

Monitoring 63

Monitoring Entertainment Hardware Latency 63

Brendan Gregg

JULY 4, 2021

This was a chance to talk about other things I've been working on, such as the present and future of hardware performance. I also wrote about these topics in detail for my recent [Systems Performance 2nd Edition] book. Note that my predictions in this talk may be wrong, but they should be thought-provoking. Ford, et al., “TCP

Performance 52

Performance Latency Hardware Storage 52

John McCalpin

DECEMBER 6, 2016

The Xeon Phi x200 (Knights Landing) has a lot of modes of operation (selected at boot time), and the latency and bandwidth characteristics are slightly different for each mode. This reduces the number of “hops” required for request/response/coherence messages on the mesh, and should reduce both latency and contention.

Latency 40

Latency Cache Testing Systems 40

Brendan Gregg

SEPTEMBER 25, 2021

A Cassandra database cluster had switched to Ubuntu and noticed write latency increased by over 30%. CLI tools The Cassandra systems were EC2 virtual machine (Xen) instances. As a Xen guest, this profile was gathered using perf(1) and the kernel's software cpu-clock soft interrupts, not the hardware NMI.

Speed 126

Speed Java AWS Virtualization 126

The Morning Paper

MARCH 17, 2020

Based on these findings, the authors present four guidelines to get the best performance out of this memory today. higher latency and lower bandwidth)… We have found the actual behavior of Optane DIMMs to be more complicated and nuanced than the "slower, persistent DRAM" label would suggest. Optane DIMMs are here!

Scalability 59

Scalability Latency Cache Media 59

HammerDB

OCTOBER 18, 2019

So lets take an Ubuntu system with Platinum 8280 CPUs with the following Ubuntu OS, reboot and check the CPU configuration before running any tests. So by default, the system boots into powersave. hardware limits: 1000 MHz - 4.00 hardware limits: 1000 MHz - 4.00 uname -a Linux ubuntu19 5.3.0-rc3-custom 3834 2696 .

Benchmarking 63

Benchmarking Performance Hardware C++ 63

Brendan Gregg

FEBRUARY 28, 2023

This talk originated from my updates to Systems Performance 2nd Edition , and this was the first time I've given this talk in person! CXL in a way allows a custom memory controller to be added to a system, to increase memory capacity, bandwidth, and overall performance. Ford, et al., “TCP

Performance 52

Performance Latency Cache Virtualization 52

Percona

MAY 2, 2023

However, it’s important to note that the optimal value may vary depending on your specific workload and hardware configuration. The answer does not consider the queue or latency of the sample, which could indicate a disk with issues. 16) and monitoring the server’s performance. The answer could be better.

Social Media 120

Social Media Database Servers Cache 120

The Morning Paper

DECEMBER 8, 2019

At the start of November I was privileged to attend HPTS (the High Performance Transaction Systems) conference in Asilomar. On the last morning of the conference Daniel Bittman presented some of the work being done in the context of the Twizzler OS project to explore new programming models for NVM.

Hardware 56

Hardware Programming Media Storage 56

Dotcom-Montior

DECEMBER 8, 2021

Systems, web applications, servers, devices, etc., SREs and DevOps teams can use these incidents to build back better and improve their systems and services. Now that we have talked about what an incident is, incident management is the process by which teams resolve these events and bring systems and services back to normal operation.

Social Media 90

Social Media Monitoring Latency DevOps 90

The Morning Paper

MAY 28, 2019

We don’t talk about energy as often as we probably should on this blog, but it’s certainly true that our data centres and various IT systems consume an awful lot of it. The goal is to produce a low-energy hardware classifier for embedded applications doing local processing of sensor data. Introducing race logic.

Energy 52

Energy Hardware Efficiency Architecture 52

Alex Russell

APRIL 29, 2021

After 20 years of neck-in-neck competition, often starting from common code lineages, there just isn't that much left to wring out of the system. Consistent improvement is the name of the game, and it can still have positive impacts, particularly as users lean on the system more heavily over time. Media Session API. Offscreen Canvas.

Media 145

Media Games Education Engineering 145

Brendan Gregg

SEPTEMBER 25, 2021

A Cassandra database cluster had switched to Ubuntu and noticed write latency increased by over 30%. CLI tools The Cassandra systems were EC2 virtual machine (Xen) instances. As a Xen guest, this profile was gathered using perf(1) and the kernel's software cpu-clock soft interrupts, not the hardware NMI.

Speed 52

Speed Java AWS Virtualization 52

Alex Russell

MARCH 6, 2021

Hardware Past As Performance Prologue. This 2GiB RAM, Android 9 stalwart features the all-too classic lines of a Quad-core A53 (1.4GHz, small mercies) CPU, tastefully presented in a charming 5.5" Sadly, data on latency is harder to get, even from Google's perch, so progress there is somewhat more difficult to judge.

Performance 124

Performance Network Cache Metrics 124

CSS - Tricks

JULY 25, 2019

Lighthouse records metrics from the browser, applies a scoring model to them, and presents an overall performance score. Estimated Input Latency. Estimated Input Latency. To successfully uncover significant differences in user experience, we suggest using a performance monitoring system (like Calibre !) In PageSpeed 5.0,

Google 71

Google Engineering Speed Mobile 71

John McCalpin

JANUARY 22, 2018

Dr. Damon McDougall gave a short presentation on this study at the IXPUG 2018 Fall Conference ( pdf ) — I originally wrote these notes to help organize my thoughts as we were preparing the IXPUG presentation, and later decided that the extra details contained here are interesting enough for me to post it.

Latency 40

Latency Hardware Code Testing 40

O'Reilly Software

APRIL 4, 2018

Fast forward to the present day and we find ourselves in a world where the number of connected devices is constantly increasing. Are there inherent time relationships in the messages that need to be preserved as they travel across the system? Clean and reprocess it? Do I need to preserve ordering? At least once? At most once?

Architecture 80

Architecture Scalability Google Operating System 80

Percona

APRIL 17, 2023

In this blog post, we will discuss the best practices on the MongoDB ecosystem applied at the Operating System (OS) and MongoDB levels. Operating System (OS) settings Swappiness Swappiness is a Linux kernel setting that influences the behavior of the Virtual Memory manager when it needs to allocate a swap, ranging from 0-100.

Best Practices 52

Best Practices Design Tuning Database 52

Adrian Cockcroft

NOVEMBER 11, 2019

A resilient system continues to operate successfully in the presence of failures. The system needs to maintain a safety margin that is capable of absorbing failure via defense in depth, and failure modes need to be prioritized to take care of the most likely and highest impact risks. The first technique is the most generally useful.

Latency 52

Latency Engineering Systems Hardware 52

Brendan Gregg

SEPTEMBER 25, 2021

A Cassandra database cluster had switched to Ubuntu and noticed write latency increased by over 30%. CLI tools The Cassandra systems were EC2 virtual machine (Xen) instances. As a Xen guest, this profile was gathered using perf(1) and the kernel's software cpu-clock soft interrupts, not the hardware NMI.

Speed 40

Speed Java AWS Virtualization 40

Adrian Cockcroft

NOVEMBER 11, 2019

A resilient system continues to operate successfully in the presence of failures. The system needs to maintain a safety margin that is capable of absorbing failure via defense in depth, and failure modes need to be prioritized to take care of the most likely and highest impact risks. The first technique is the most generally useful.

Latency 53

Latency Engineering Systems Hardware 53

SQL Server According to Bob

JANUARY 11, 2020

This will help you increase system performance and avoid I/O environment errors. Important When planning, deploying, and maintaining a Microsoft SQL Server installation, ensure that the I/O system supports all the factors outlined in this article. This White Paper is for informational purposes only.

Servers 40

Servers Cache Media Hardware 40

HammerDB

OCTOBER 12, 2018

As with the previous guides as an Intel employee (#IAMINTEL) the examples are taken from a MySQL 8 on Linux on Intel system and the approach is the same for whatever system you are testing although some of the settings you see may be different. For this reason sysbench presents a vastly more simplified workload than HammerDB. .

Best Practices 40

Best Practices Scalability Performance C++ 40

Alex Russell

JULY 14, 2021

The predominant desktop situation is relatively straightforward: Browsers handle links, and non-browsers defer loading http and https URLs to the system, which in turn invokes the default browser. Instead, they load content referred by their hosting native app in system-provided WebViews. The Baseline Scenario #.

Google 121

Google Mobile Engineering Internet 121

Alex Russell

JANUARY 30, 2024

HTML, CSS, images, and fonts can all be parsed and run at near wire speeds on low-end hardware, but JavaScript is at least three times more expensive, byte-for-byte. If you or your company are able to generate a credible worldwide latency estimate in the higher percentiles for next year's update, please get in touch.

Performance 90

Performance Network Mobile Speed 90

Alex Russell

JUNE 20, 2020

Recall that browsers arose atop platforms that universally provided a host of services by the late 90's: system fonts (to support layout programs, including WYSIWYG editing). In each quantum of time, hardware and OS vendors press ahead, adding features. This is often determined by hardware integration and device replacement rates.

Hardware 55

Hardware Games Java Media 55

The Morning Paper

SEPTEMBER 12, 2019

If you think about a system like Procella that’s combining transactional and analytic workloads on top of a cloud-native architecture, extensions to SQL for streaming , dataflow based materialized views (see e.g. It would be a bold departure from the current imperative high-level DSL based approach to deep learning taken by TensorFlow et al.,

Network 44

Network Database Programming Hardware 44

HammerDB

DECEMBER 5, 2020

A full understanding of why this is important requires some knowledge of the evolution of database hardware and software. For TPC-C this meant enough available spindles to reduce I/O latency and for TPC-H enough bandwidth for data throughput. I.e. if system A generated 1.5X

C++ 40

C++ Benchmarking Virtualization Open Source 40

SQL Server According to Bob

JANUARY 11, 2020

​​ Because Microsoft must respond to changing market conditions, it should not be interpreted to be a commitment on the part of Microsoft, and Microsoft cannot guarantee the accuracy of any information presented after the date of publication. This White Paper is for informational purposes only.

Servers 40

Servers Cache Database Media 40