Brendan Gregg

FEBRUARY 28, 2023

At USENIX SREcon22 APAC I gave the opening keynote on the future of computer performance, rounding up the latest developments and making predictions of where I see things heading. This talk originated from my updates to Systems Performance 2nd Edition , and this was the first time I've given this talk in person!

Performance 52

Performance Latency Cache Virtualization 52

Smashing Magazine

DECEMBER 2, 2021

The project started in 2012 with the idea of providing technical content during the Christmas Advent period, so keep looking for nice things under the Java Christmas tree! ??. Good ol’ PerfPlanet is back for another season with all things speed and web performance. The Java Advent 2021 is here! JavaAdvent on Twitter.

Design 107

Design Development Programming Website 107

All Things Distributed

JULY 19, 2012

High Performance I/O Instances for Amazon EC2. By Werner Vogels on 19 July 2012 12:00 AM. These include the likes of high performance computation, for which we introduced the Cluster Compute and Cluster GPU instance types. Databases are one particular area that for scaling can benefit tremendously from high performance I/O.

Cloud 129

Cloud Performance AWS Virtualization 129

Dynatrace

NOVEMBER 16, 2021

The dashboard was built during my recent Performance Clinic on k8s monitoring at Scale with Prometheus and Dynatrace. Since its launch in 2012, Prometheus has become the standard technology to collect metrics in a Kubernetes cluster. Here we’ll explore how to collect Prometheus metrics and what you can achieve with them.

Metrics 206

Metrics Infrastructure Open Source Database 206

High Scalability

MARCH 22, 2019

BrentToderian : What city went from 14% of all trips by bike in 2001, to 22% by 2012, then leaped to 30% in 3 years by 2015, & 35% by 2018? What happens when no new open source comes out of the smaller companies, and the big-3 decide they don't really need or want to play nice anymore? 3B this year, $4B next year. Debt skyrocketing.

Internet 134

Internet Internet Scalability Wireless 134

Dynatrace

DECEMBER 18, 2023

Named after the Greek god who brought fire down from Mount Olympus, Prometheus metrics have been transforming observability since the project’s inception in 2012. The batch processor, to improve performance by batching requests. The OTLP/HTTP exporter, to perform the data export to Dynatrace.

Metrics 182

Metrics Engineering Energy Tuning 182

Dynatrace

APRIL 2, 2020

Performance has always mattered. With sniffers, network performance engineers were ultimately able to relate wire data performance to what business owners expected from the networks they financed. The network performance engineer who managed the sniffer was the guru. The DC RUM wave. Technology developments come in waves.

Network 166

Network Traffic Monitoring Java 166

Strategic Tech

FEBRUARY 27, 2021

In February 2012 I began working for a new company. Every step in a developer’s journey is an opportunity to improve their experience and improve the performance of the organization. Back in 2012, my CTO was passionate about “every developer pushing [meaningful] code to production on their first day”. I was lost for words.

Development 98

Development Strategy Infrastructure Metrics 98

SQL Performance

MAY 16, 2019

Achieving minimal logging using INSERT.SELECT into an empty clustered index target is not quite as simple as described in the Data Performance Loading Guide. The Data Loading Performance Guide contains a high-level summary of the conditions required for minimal logging into clustered tables: This post is concerned with the top row only.

Cache 67

Cache Servers Testing Storage 67

All Things Distributed

AUGUST 17, 2012

By Werner Vogels on 17 August 2012 07:00 PM. I am in São Paolo, Brazil for the 2012 AWS Latin America Summit and for The Next Web Latin America conference. Matt Welsh s thesis work at Berkeley was on building high-performance internet services. Werner Vogels weblog on building scalable and robust distributed systems.

Architecture 98

Architecture AWS Internet Internet 98

Tim Kadlec

JUNE 21, 2022

HTTP/3 work started in 2012 with Google working on QUIC, adopted by IETF in 2017, RFC’s published in June 2022. Testing the Performance of HTTP/3. Protocol performance is rarely low0hanging fruit. At #SmashingConf San Francisco , Robin Marx gave a presentation about HTTP/3. Here are my notes. What and Why?

Network 64

Network Mobile Google Cache 64

All Things Distributed

AUGUST 23, 2013

I want to kick off the fall readings not with an historical paper but with two that detail GraphLab , an excellent framework for high performance machine learning that originally has been built by the Carlos Guestrin. The research in the papers has been performed on Amazon EC2. 716-727 (2012).

Cloud 83

Cloud Operating System Design Systems 83

Randon ASCII

OCTOBER 1, 2023

Investigating weird Chrome performance problems is literally my job so we started chatting. The difference is dramatic: The performance of scanning the memory in the normal mode is quite consistent, taking about 0.2 The answer is later in this post. Then it does the same thing with the process set to background mode. ms per scan.

Efficiency 98

Efficiency Processing Programming Systems 98

Randon ASCII

JANUARY 17, 2023

In general if you want to understand why your computer is performing badly you need to record and analyze a trace. Raymond Chen gives several reasons for why Windows Error Reporting doesn’t restart crashed processes before uploading the report (circa 2012) but I don’t find those reasons entirely compelling, especially in the start-menu case.

Programming 114

Programming Processing Code Performance 114

SQL Performance

OCTOBER 9, 2019

The EOMONTH function was introduced in SQL Server 2012. SQL Server 2012 introduced the function CONCAT. Similar to the EOMONTH function, many people don't realize that starting already with SQL Server 2012, the LOG function supports a second parameter that allows you to indicate the logarithm's base. Open the cursor.

C++ 111

C++ Database Servers Code 111

SQL Performance

MARCH 15, 2021

This wasn't for performance reasons, directly, or to simply follow Microsoft's documented best practices, but revolved more around the decisions you might make when you only have access to some of the data. In my last post , I illustrated one reason you should stop using deprecated system tables like sysprocesses.

Servers 64

Servers Azure Best Practices Monitoring 64

All Things Distributed

NOVEMBER 13, 2018

In 2012, Amazon opened its first Italian office and its first Italian point of presence (PoP) based in Milan. Lamborghini, the world-famous manufacturer of elite, luxury sports cars based in Italy, has been using AWS to reduce the cost of their infrastructure by 50 percent, while also achieving better performance and scalability.

AWS 167

AWS Energy Automotive Traffic 167

SQL Performance

FEBRUARY 22, 2021

An excuse I used to hear a lot was that sys.dm_exec_sessions was missing open_tran , but open_transaction_count was added back in SQL Server 2012. So there is very little reason to even think about using sys.sysprocesses today. dm_os_performance_counters.

Latency 57

Latency Servers Monitoring Systems 57

John McCalpin

JANUARY 9, 2019

This week I updated three of those posts — two really old ones (primarily of interest to computer architecture historians), and one from 2018: July 2012: Local and Remote Memory Latency on AMD Processors in 2-socket and 4-socket servers. December 2013: Notes on Memory Bandwidth on the Xeon Phi (Knights Corner) Coprocessor.

Latency 40

Latency Architecture Servers Performance 40

All Things Distributed

NOVEMBER 17, 2012

By Werner Vogels on 17 November 2012 11:00 PM. Scale and performance in a distributed file system , John H. All Things Distributed. Werner Vogels weblog on building scalable and robust distributed systems. Back-to-Basics Weekend Reading - The Andrew File System. Comments (). Howard, Michael L. Kazar, Sherri G. Menees, David A.

Systems 88

Systems Government Cache Scalability 88

SQL Server According to Bob

JUNE 6, 2016

As indicated, SQL Server 2016 added the use of SSE and AVX instructions in various code paths to improve performance. Windows 2012 R2 and newer releases have been enlightened the Crypto* APIs. Version of SQL Server 2012, 2014, and 2016 using the Crypto* APIs take advantage of the hardware implementations through the API support.

Servers 40

Servers Hardware Software Engineering Software 40

All Things Distributed

MAY 8, 2012

By Werner Vogels on 08 May 2012 02:00 PM. Amazon RDS manages all these time consuming database administration tasks including patch management, striping the storage for better performance, and database and log backups for disaster recovery, enabling developers to focus more on their applications. All Things Distributed. Comments ().

Cloud 116

Cloud Development AWS Servers 116

All Things Distributed

FEBRUARY 4, 2014

In 2012,AWS announced 159 new features and services. AWS Trusted Advisor monitors customer configurations, compares them to known best practices, and then notifies customers where opportunities exist to improve performance, enhance security, or save money. Yes,we are actively telling customers they’re paying us more than they need to.

AWS 134

AWS Energy Innovation Best Practices 134

Tim Kadlec

DECEMBER 29, 2014

“RWD Is Bad for Performance” Is Good for Performance. I’ve found the “responsive design is bad for performance” myth to be really good for performance. Too often we focus merely on size and request count when discussing the use of JavaScript from a performance perspective. Other years.

Metrics 49

Metrics Design Performance 49

All Things Distributed

DECEMBER 27, 2012

By Werner Vogels on 27 December 2012 12:00 PM. By storing your website in Amazon S3, you suddenly no longer have to worry about scaling, replication, performance, security, etc. Werner Vogels weblog on building scalable and robust distributed systems. My Best Christmas Present â?? Root Domain Support for Amazon S3 Website Hosting.

Website 127

Website Traffic Storage Scalability 127

ACM Sigarch

APRIL 26, 2018

TFLOPS deep learning performance, and is equipped with 32GB of high-bandwidth memory (HBM) capacity. The new DGX-2, which achieves 2 petaFLOPS in the system, combines 16 fully interconnected GPUS with 10x the deep learning performance. The new GV100 packs 7.4 TFLOPS FP-64, 14.8 TFLOPS FP-32 and 118.5

Hardware 99

Hardware Artificial Intelligence Architecture Google 99

SQL Server According to Bob

JULY 12, 2016

Trace flags 2562 and 2549 are documented in knowledgebase article: [link] and the blog post [link] highlights the SQL 2016 DBCC performance improvements. For SQL Server 2008, 2012 and 2014 the trace flag behavior remains the same.

Servers 40

Servers Design Engineering Database 40

Tim Kadlec

DECEMBER 29, 2015

It cripples performance, limits reach and reduces stability. Choosing Performance. Performance is a decision. 2012 Favorites. If your client-side MVC framework does not support server-side rendering, that is a bug. Taking Let’s Encrypt for a Spin. Unfortunately, moving to HTTPS can be kind of painful. Other years.

Energy 40

Energy Google Servers Processing 40

All Things Distributed

SEPTEMBER 25, 2012

By Werner Vogels on 25 September 2012 06:00 PM. Following the huge success of being able to provision a consistent, user-requested I/O rate for DynamoDB and Elastic Block Store (EBS), the AWS Database Services team has now released Provisioned IOPS, a new high performance storage option for the Amazon Relational Database Service (Amazon RDS).

Cloud 118

Cloud AWS Storage Database 118

Sutter's Mill

JUNE 29, 2012

Here’s another deep session for C&B 2012 on August 5-8 – if you haven’t registered yet, register soon. But there’s just so much about C++11 to cover, so here’s a fourth brand-new session I’ll do at C&B 2012 that goes deeper on its topic than I’ve ever been willing to go before. This session in one word: Deep. It’ll be a blast.

Hardware 40

Hardware C++ Code Design 40

Smashing Magazine

JUNE 30, 2021

The Problem Of Unpredictable Performance. At the same time, we have a widening performance gap. In terms for performance, these phones have plateaued at the performance of a 2012 iPhone. Applications built for the Web are expected to run on devices that fall anywhere on this broad performance spectrum.

Games 145

Games Architecture Code C++ 145

All Things Distributed

FEBRUARY 22, 2012

By Werner Vogels on 22 February 2012 12:01 AM. Often an application consists of several different tasks to be performed in particular sequence driven by a set of dynamic conditions. Developers implement workers to perform tasks. Comments (). Amazon SWF is an orchestration service for building scalable distributed applications.

Cloud 130

Cloud AWS Java Scalability 130

Speed Curve

DECEMBER 19, 2018

Understanding the importance of loading scripts asynchronously might help increase adoption of this critical performance improvement, so we're going to walk through the evolution of async script loading starting way back in 2007. The preloader is the most important performance improvement browsers ever made. Finally: Async.

Internet 53

Internet Internet Website Traffic 53

All Things Distributed

NOVEMBER 8, 2012

By Werner Vogels on 08 November 2012 03:00 PM. Customers from various verticals (media, social gaming, mobile, news, advertisement) such as Netflix, Shazam and Scopely have used SQS in variety of use-cases requiring loose coupling and high performance. All Things Distributed. Improving the Cloud - More Efficient Queuing with SQS.

Efficiency 107

Efficiency Cloud Games Scalability 107

Speed Curve

AUGUST 8, 2017

I've written a lot of posts about page bloat, starting way back in 2012, when the average page hit 1MB. Because the average page is now 3MB in size, and this seems like a good time to pause, check our assumptions, and ask ourselves: Is there any reason to care about page size as a performance metric? The answer was no.

Metrics 81

Metrics Website Speed Benchmarking 81

Tim Kadlec

DECEMBER 26, 2013

Time for the long-running (since way back in 2012 !) Setting a Performance Budget. Getting people invested in performance has to happen from the start of a project. Setting a performance budget is a great way to help with that. annual tradition of listing the top five most popular posts on this site of the past year.

Media 40

Media Design Technology Technology 40

SQL Server According to Bob

JUNE 3, 2016

. ‘It Just Runs Faster’ – Out of the box SQL Server 2016 DBCC provides you better performance, scale while shrinking your maintenance window(s.). The DBCC CheckDB demonstration loads a table and demonstrates the performance improvement. Ryan Stonecipher – Principle SQL Server Software Engineer. or newer release.

Hardware 40

Hardware Cache Servers Software Engineering 40