Five Steps to Solving Software Performance Problems
Hand-out accompanying a seminar on performance testing. Fairly good core rundown of what needs to be done by a Performance Engineer when covering a project.
Jeffrey Kelly
http://sailwhatcom.com
Performance engineering encompasses the set of roles, skills, activities, practices, tools, and deliverables applied at every phase of the Systems Development Lifecycle which ensures that a solution will be designed, implemented, and operationally supported to meet the non-functional requirements. (Wikipedia)
Wednesday, December 26, 2007
Monday, December 17, 2007
System-performance modeling for massively multiplayer online role-playing games
The title pretty much sums it up. You will need to create a free ZDNet account if you do not have one.
Jeffrey Kelly
http://sailwhatcom.com
The title pretty much sums it up. You will need to create a free ZDNet account if you do not have one.
Jeffrey Kelly
http://sailwhatcom.com
Monday, December 10, 2007
The Tweaking Experience
Ever wondered how to tweak the Windows UI from the registry? How about how to change the frequency that preload information is cleared? What about defragmenting your boot file? This help-like web site has it all.
Jeffrey Kelly
http://sailwhatcom.com
Ever wondered how to tweak the Windows UI from the registry? How about how to change the frequency that preload information is cleared? What about defragmenting your boot file? This help-like web site has it all.
Jeffrey Kelly
http://sailwhatcom.com
Monday, December 03, 2007
Boost AJAX Performance Using Local Storage
Article about Ajax performance and how to help it out on the client side.
Jeffrey Kelly
http://sailwhatcom.com
Article about Ajax performance and how to help it out on the client side.
Jeffrey Kelly
http://sailwhatcom.com
Thursday, November 29, 2007
Beyond performance testing
Extensive article on what comes after Performance Testing... Performance Engineering. Seems quite appropriate based on the changes the Stress team is going through.
Jeffrey Kelly
http://sailwhatcom.com
Extensive article on what comes after Performance Testing... Performance Engineering. Seems quite appropriate based on the changes the Stress team is going through.
Jeffrey Kelly
http://sailwhatcom.com
Monday, November 19, 2007
Ten ways to improve testing, performance of Web 2.0 applications
Interesting & brief article on Web 2.0, what it is, a gross generalization about customer experience expectations and some things companies can do to score better.
Jeffrey Kelly
http://sailwhatcom.com
Interesting & brief article on Web 2.0, what it is, a gross generalization about customer experience expectations and some things companies can do to score better.
Jeffrey Kelly
http://sailwhatcom.com
Monday, November 12, 2007
An Explanation of Performance Testing on an Agile Team (Part 1 of 2)
Part 1 of 2 articles about performance testing in an Agile development team. Quite in depth...
Jeffrey Kelly
http://sailwhatcom.com
Part 1 of 2 articles about performance testing in an Agile development team. Quite in depth...
Jeffrey Kelly
http://sailwhatcom.com
Monday, November 05, 2007
Space Station: Internal NASA Reports Explain Origins of June Computer Crisis
Article covering the event and troubleshooting by geographically dispersed and culturally different teams. Performance testers find these sorts of problems and "solutions" all the time when working across multiple project teams or even the same team with different disciplines (Dev, DB Dev, Tester).
Jeffrey Kelly
http://sailwhatcom.com
Article covering the event and troubleshooting by geographically dispersed and culturally different teams. Performance testers find these sorts of problems and "solutions" all the time when working across multiple project teams or even the same team with different disciplines (Dev, DB Dev, Tester).
Jeffrey Kelly
http://sailwhatcom.com
Monday, October 08, 2007
Performance Testing 101
Here is an article about a different kind of performance testing and I think you will see many parallels with what we do today in the software industry.
Jeffrey Kelly
http://sailwhatcom.com
Here is an article about a different kind of performance testing and I think you will see many parallels with what we do today in the software industry.
Jeffrey Kelly
http://sailwhatcom.com
Monday, October 01, 2007
Performance Enhancements in the Next Generation TCP/IP Stack
A whole list of changes are coming to the TCP/IP stack in Vista and Server 2008 products. This article covers some of the performance impacting ones... generally meant to improve performance.
Jeffrey Kelly
http://sailwhatcom.com
A whole list of changes are coming to the TCP/IP stack in Vista and Server 2008 products. This article covers some of the performance impacting ones... generally meant to improve performance.
Jeffrey Kelly
http://sailwhatcom.com
Monday, September 24, 2007
The Product is the Message; the Message is the Product
Interesting article from John Maeda (who's blog has previously been the link of the week) that forces contemplation of hardware, software and which is more important. This link is not directly performance related, however when I read it I immediately thought about the testing, reports and website that my team maintains and what role each part plays, or should play in communication with our customers. And hence, here it is in your inbox.
Jeffrey Kelly
http://sailwhatcom.com
Interesting article from John Maeda (who's blog has previously been the link of the week) that forces contemplation of hardware, software and which is more important. This link is not directly performance related, however when I read it I immediately thought about the testing, reports and website that my team maintains and what role each part plays, or should play in communication with our customers. And hence, here it is in your inbox.
Jeffrey Kelly
http://sailwhatcom.com
Wednesday, September 19, 2007
Ever wondered if you had a DNS performance problem? Here is a little app that may be able to help.
http://www.codeproject.com/cs/internet/DNSTester.asp
Jeffrey Kelly
http://sailwhatcom.com
http://www.codeproject.com/cs/internet/DNSTester.asp
Jeffrey Kelly
http://sailwhatcom.com
Tuesday, September 11, 2007
CNET Labs - How We Test
Ever wondered how CNET tests hardware? Here is some information along this line...
Jeffrey Kelly
http://sailwhatcom.com
Ever wondered how CNET tests hardware? Here is some information along this line...
Jeffrey Kelly
http://sailwhatcom.com
Wednesday, August 22, 2007
CNET 2008 antivirus performance test scores
Check out this site from CNET on how they performance test antivirus software and pick their Editor's Choice. Interesting read.
Jeffrey Kelly
http://sailwhatcom.com
Check out this site from CNET on how they performance test antivirus software and pick their Editor's Choice. Interesting read.
Jeffrey Kelly
http://sailwhatcom.com
Wednesday, August 15, 2007
Web Services Performance: Comparing J2EE and .NET Framework
This whitepaper documents a study comparing J2EE and .NET performance for web services. In all cases, this stufy found that J2EE outperformed .NET. There is a good description of the test, the test bed, tood configuration, and the tuning implemented.
Jeffrey Kelly
http://sailwhatcom.com
This whitepaper documents a study comparing J2EE and .NET performance for web services. In all cases, this stufy found that J2EE outperformed .NET. There is a good description of the test, the test bed, tood configuration, and the tuning implemented.
Jeffrey Kelly
http://sailwhatcom.com
Monday, August 13, 2007
Beijing starts IT testing for Olympics
China is excited about the Olympics and testing for the major IT systems has begun. Read on for more.
Jeffrey Kelly
http://sailwhatcom.com
China is excited about the Olympics and testing for the major IT systems has begun. Read on for more.
Jeffrey Kelly
http://sailwhatcom.com
Monday, August 06, 2007
BCCHardware: Core 2 Duo vs Core 2 Quad
This review compares performance from Intel's new Core 2 Duo and Core 2 Quad processors. There is some extensive benchmarking results and some overclocking. There is also some good matter-of-fact recommendations if you are purchasing a new processor.
Jeffrey Kelly
http://sailwhatcom.com
This review compares performance from Intel's new Core 2 Duo and Core 2 Quad processors. There is some extensive benchmarking results and some overclocking. There is also some good matter-of-fact recommendations if you are purchasing a new processor.
Jeffrey Kelly
http://sailwhatcom.com
Monday, July 30, 2007
What is OGSA-DQP?
This link is the kind that you will need some uninterrupted brain time to look at. This is an implementation of DQP (Distributed Query Processing) which is a way to abstract queries for data across multiple componsnts (databases, data processors, formatters, etc) in a distributed way.
Take some time to dig in and contemplate how some of our Expedia systems would benefit or suffer from such a model.
Jeffrey Kelly
http://sailwhatcom.com
This link is the kind that you will need some uninterrupted brain time to look at. This is an implementation of DQP (Distributed Query Processing) which is a way to abstract queries for data across multiple componsnts (databases, data processors, formatters, etc) in a distributed way.
Take some time to dig in and contemplate how some of our Expedia systems would benefit or suffer from such a model.
Jeffrey Kelly
http://sailwhatcom.com
Tuesday, July 24, 2007
5 Stress Reduction Tips
Quick article with some Stress reduction techniques.
Jeffrey Kelly
http://sailwhatcom.com
Quick article with some Stress reduction techniques.
Jeffrey Kelly
http://sailwhatcom.com
Monday, July 16, 2007
Online travel sites court unmanaged business users
Interesting article about business travelers who are not part of a company/firm that contracts travel services and what OLTPs are doing to try to bring them in as customers.
Jeffrey Kelly
http://sailwhatcom.com
Interesting article about business travelers who are not part of a company/firm that contracts travel services and what OLTPs are doing to try to bring them in as customers.
Jeffrey Kelly
http://sailwhatcom.com
Monday, July 09, 2007
Troubleshooting Performance Problems in SQL Server 2005
This TechNet article covers DB Administrators with basic information on troubleshooting SQL Server performance slow downs. Somce of the topics include Bottlenecks, TempDB, and Query performance.
Jeffrey Kelly
http://sailwhatcom.com
This TechNet article covers DB Administrators with basic information on troubleshooting SQL Server performance slow downs. Somce of the topics include Bottlenecks, TempDB, and Query performance.
Jeffrey Kelly
http://sailwhatcom.com
Monday, June 04, 2007
How to use the SQLIOSim utility to simulate SQL Server activity on a disk subsystem
SQLIOSim is the latest version of SQLIOStress which are used to test the Disk Subsystem behind a SQL Server implementation. This document contains an overview of the tool, directions for execution, and some error handling information. Cool tool if you need to test this sort of thing!
Jeffrey Kelly
http://sailwhatcom.com
SQLIOSim is the latest version of SQLIOStress which are used to test the Disk Subsystem behind a SQL Server implementation. This document contains an overview of the tool, directions for execution, and some error handling information. Cool tool if you need to test this sort of thing!
Jeffrey Kelly
http://sailwhatcom.com
Tuesday, May 29, 2007
E-Mail Functionality in SQL Server 2005
This article covers some of the reasons why email in SQL 2000 was not a realistic piece of functionality for production use, as well as gives directions on how to install, configure, and use Database Mail (the SQL 2k5 version).
Jeffrey Kelly
http://sailwhatcom.com
This article covers some of the reasons why email in SQL 2000 was not a realistic piece of functionality for production use, as well as gives directions on how to install, configure, and use Database Mail (the SQL 2k5 version).
Jeffrey Kelly
http://sailwhatcom.com
Monday, May 21, 2007
Designing Performance Testing Metrics into Highly Distributed J2EE Applications
This article covers one case study on including performance data collection in a vastly distributed system. Come to think of it, Expedia.com is a vastly distributed system...
Jeffrey Kelly
http://sailwhatcom.com
This article covers one case study on including performance data collection in a vastly distributed system. Come to think of it, Expedia.com is a vastly distributed system...
Jeffrey Kelly
http://sailwhatcom.com
Monday, May 14, 2007
The Last Mile
This PDF defines the "Last Mile" in a performance test context. In short, it is a method to find a good representation of what end users of a website experience when they interact with a site.
Jeffrey Kelly
http://sailwhatcom.com
This PDF defines the "Last Mile" in a performance test context. In short, it is a method to find a good representation of what end users of a website experience when they interact with a site.
Jeffrey Kelly
http://sailwhatcom.com
Monday, May 07, 2007
Windows SysInternals
If it has been a while since you looked into the SysInternals utilities available, you should check out this site. You will notice that the URL is now Microsoft... my how times change.
Jeffrey Kelly
http://sailwhatcom.com
If it has been a while since you looked into the SysInternals utilities available, you should check out this site. You will notice that the URL is now Microsoft... my how times change.
Jeffrey Kelly
http://sailwhatcom.com
Monday, April 30, 2007
How We Test - Vehicle Performance Testing
Today launches a mini-series into performance testing in other industries. Today's link provides a view into performance testing sports cars.
Jeffrey Kelly
http://sailwhatcom.com
Today launches a mini-series into performance testing in other industries. Today's link provides a view into performance testing sports cars.
Jeffrey Kelly
http://sailwhatcom.com
Monday, April 23, 2007
Custom Google Perf Test Search Engine
Excerpt: "Are you tired of all the results for tuning sports cars, training for a marathon and measuring employee productivity when searching for material related to software performance testing? We were. That is why PerfTestPlus customized our own Google search engine. We found it so useful that we decided to make it available for you to use as well. We hope it makes your searching more productive."
Comments: Scott Barber is an independent contractor/owner of PerfTestPlus, Inc. He is a frequent speaker at conferences and has many years of performance test experience. I encourage you to also check out his blog on TestingReflections.com.
Jeffrey Kelly
http://sailwhatcom.com
Excerpt: "Are you tired of all the results for tuning sports cars, training for a marathon and measuring employee productivity when searching for material related to software performance testing? We were. That is why PerfTestPlus customized our own Google search engine. We found it so useful that we decided to make it available for you to use as well. We hope it makes your searching more productive."
Comments: Scott Barber is an independent contractor/owner of PerfTestPlus, Inc. He is a frequent speaker at conferences and has many years of performance test experience. I encourage you to also check out his blog on TestingReflections.com.
Jeffrey Kelly
http://sailwhatcom.com
Monday, April 16, 2007
Creating an Effective Test Plan
An article from Parasoft on the Who, What, When, Where, Why, How of effective test planning. Quick read.
Jeffrey Kelly
http://sailwhatcom.com
An article from Parasoft on the Who, What, When, Where, Why, How of effective test planning. Quick read.
Jeffrey Kelly
http://sailwhatcom.com
Tuesday, April 10, 2007
A World-Class Infrastructure for Performance Testing
Article presented at STARWEST in 2002 which is a SQE conference about building out a performance test world. Interesting content though slightly dated.
Jeffrey Kelly
http://sailwhatcom.com
Article presented at STARWEST in 2002 which is a SQE conference about building out a performance test world. Interesting content though slightly dated.
Jeffrey Kelly
http://sailwhatcom.com
Monday, April 02, 2007
Here are some links on how to make use of performance counters in .NET code:
http://msdn2.microsoft.com/en-us/library/ms979204.aspx
http://www.codeproject.com/dotnet/perfcounter.asp
http://www.codeproject.com/vb/net/PerformanceCounterUtils.asp
http://www.codeproject.com/system/cpuusage.asp
Jeffrey Kelly
http://sailwhatcom.com
http://msdn2.microsoft.com/en-us/library/ms979204.aspx
http://www.codeproject.com/dotnet/perfcounter.asp
http://www.codeproject.com/vb/net/PerformanceCounterUtils.asp
http://www.codeproject.com/system/cpuusage.asp
Jeffrey Kelly
http://sailwhatcom.com
Monday, March 26, 2007
Top Ten Readings In Software Performance Engineering
Performance engineering is an area the Perf Test team will be breaking in to shortly so these make excellent primers. This link reveals a slightly dated list of readings on the subject.
Jeffrey Kelly
http://sailwhatcom.com
Performance engineering is an area the Perf Test team will be breaking in to shortly so these make excellent primers. This link reveals a slightly dated list of readings on the subject.
Jeffrey Kelly
http://sailwhatcom.com
Monday, March 19, 2007
Measuring SOA Performance Is a Complex Art
A fairly thin article on SOA and how managing and monitoring is a new beast from traditional IT software systems.
Jeffrey Kelly
http://sailwhatcom.com
A fairly thin article on SOA and how managing and monitoring is a new beast from traditional IT software systems.
Jeffrey Kelly
http://sailwhatcom.com
Monday, March 12, 2007
Internet Protocol version 6 (IPv6) Conformance and Performance Testing
This is quite a long white paper on IPv6 and the transition from v4 to v6 and the testing considerations that should cross your mind. Information in the article includes Background, Benefits, Operation, Testing and the approach/tools used by this vendor.
Jeffrey Kelly
http://sailwhatcom.com
This is quite a long white paper on IPv6 and the transition from v4 to v6 and the testing considerations that should cross your mind. Information in the article includes Background, Benefits, Operation, Testing and the approach/tools used by this vendor.
Jeffrey Kelly
http://sailwhatcom.com
Monday, March 05, 2007
Vista Performance Shootout: Upgrade Vs. Clean Installation
Here is an interesting article on the new deployment process of Windows Vista and how it compares (performance wise) when running an upgrade and a clean install.
Jeffrey Kelly
http://sailwhatcom.com
Here is an interesting article on the new deployment process of Windows Vista and how it compares (performance wise) when running an upgrade and a clean install.
Jeffrey Kelly
http://sailwhatcom.com
Monday, February 26, 2007
The Browser is Slow
This is a look at the render time and browser data-handling capability of IE7. There are some interesting comparison charts between Safari, Firefox, IE6 and IE7. This was written back when IE7 was in Beta 1 so some things may have changed, but major changes are unlikely!
Jeffrey Kelly
http://sailwhatcom.com
This is a look at the render time and browser data-handling capability of IE7. There are some interesting comparison charts between Safari, Firefox, IE6 and IE7. This was written back when IE7 was in Beta 1 so some things may have changed, but major changes are unlikely!
Jeffrey Kelly
http://sailwhatcom.com
Monday, February 19, 2007
Gordon's car fails post-race inspection
This is an article about NASCAR which, you may know, started the 2007 season on Sunday with the Daytona 500. This race is arguably the toughest and most popular race of the entire season. Jeff Gordon, a popular and very successful driver was driving a car that didn't conform to the performance requirements which resulted in an unfair advantage for the race. Mr. Gordon was penalized, along with several other drivers who committed other violations. This brings us to an interesting conversation on the two types of performance requirements.
At Expedia we test systems every day that struggle to meet the performance requirements set by business owners, project managers, and test teams. What we must remember is that there are disciplines where performance is so good, that performance needs to be restricted. Wouldn't it be great if we were one of those systems? Our customers would call in saying that the homepage loaded too fast, or that they switched to dial-up because they couldn't handle the speed of our flight searches. Our partners would call in saying we need to slow down our support systems because they were too quick for the system on the partner's side of the connection. What a different world it would be in Performance Test land if this were the case!
My hope is that next time you consider granting an exception to an Expedia-style performance requirement, you think of the NASCAR style and suggest that the project team give better performance one more try before they throw-in-the-towel.
Jeffrey Kelly
http://sailwhatcom.com
This is an article about NASCAR which, you may know, started the 2007 season on Sunday with the Daytona 500. This race is arguably the toughest and most popular race of the entire season. Jeff Gordon, a popular and very successful driver was driving a car that didn't conform to the performance requirements which resulted in an unfair advantage for the race. Mr. Gordon was penalized, along with several other drivers who committed other violations. This brings us to an interesting conversation on the two types of performance requirements.
At Expedia we test systems every day that struggle to meet the performance requirements set by business owners, project managers, and test teams. What we must remember is that there are disciplines where performance is so good, that performance needs to be restricted. Wouldn't it be great if we were one of those systems? Our customers would call in saying that the homepage loaded too fast, or that they switched to dial-up because they couldn't handle the speed of our flight searches. Our partners would call in saying we need to slow down our support systems because they were too quick for the system on the partner's side of the connection. What a different world it would be in Performance Test land if this were the case!
My hope is that next time you consider granting an exception to an Expedia-style performance requirement, you think of the NASCAR style and suggest that the project team give better performance one more try before they throw-in-the-towel.
Jeffrey Kelly
http://sailwhatcom.com
Monday, February 12, 2007
Again about performance requirements (Testing Reflections)
Interesting comments about performance requirements versus recommendations as well as links to other related posts. We maintain recommendations for Expedia pages on the performance test wiki under the project consultation section.
Jeffrey Kelly
http://sailwhatcom.com
Interesting comments about performance requirements versus recommendations as well as links to other related posts. We maintain recommendations for Expedia pages on the performance test wiki under the project consultation section.
Jeffrey Kelly
http://sailwhatcom.com
Monday, February 05, 2007
Here are a couple links on Apache Tomcat performance both on the tuning side and on the benchmark side. As we're moving this direction with E3 R1, it is right up our alley.
Tomcat Tuning Notes
Comparing Apache Tomcat Performance Across Platforms
Jeffrey Kelly
http://sailwhatcom.com
Tomcat Tuning Notes
Comparing Apache Tomcat Performance Across Platforms
Jeffrey Kelly
http://sailwhatcom.com
Monday, January 29, 2007
Boost Tomcat performance for static content
This article is a good overview of how to speed up web server performance. In this case the software is Apache Tomcat but most of the content can be generalized to any web server suite.
Jeffrey Kelly
http://sailwhatcom.com
This article is a good overview of how to speed up web server performance. In this case the software is Apache Tomcat but most of the content can be generalized to any web server suite.
Jeffrey Kelly
http://sailwhatcom.com
Monday, January 15, 2007
Here are some links on Windows Vista performance. Most are more consumer-oriented than business oriented, but I'm sure as critical readers you'll plow through the fluff.
TechNet: Windows Vista: Reliability and Performance
Why Windows Vista Won't Suck
Windows Vista Performance Enhancements
Putting Vista in The Fast Lane
Jeffrey Kelly
http://sailwhatcom.com
TechNet: Windows Vista: Reliability and Performance
Why Windows Vista Won't Suck
Windows Vista Performance Enhancements
Putting Vista in The Fast Lane
Jeffrey Kelly
http://sailwhatcom.com
Monday, January 08, 2007
Performance Testing Web Services: Strategies & Best Practices
This add-heavy article covers web service testing from "Test Planning" to "What Are We Measuring" including some code examples of problems and some considerations for how to test and what load to use.
Jeffrey Kelly
http://sailwhatcom.com
This add-heavy article covers web service testing from "Test Planning" to "What Are We Measuring" including some code examples of problems and some considerations for how to test and what load to use.
Jeffrey Kelly
http://sailwhatcom.com
Subscribe to:
Comments (Atom)