Posted: January 19th, 2010 | Author: TnT Admin | Filed under: Planning | Tags: Bottleneck, hardware, installation, Load Testing, LoadRunner, Planning, probe | No Comments »
Seldom, we discuss about the planning and implementation of the load test in our site. As such, we are taking a change in this and will be touching on some considerations you as a performance tester/test manager/consultant should be thinking for every load test. These considerations enable you to smooth up the process of the entire load test and reduce any load test risks associated with it. More than often, we focus on gathering the customers requirements for the load test such as user amount to generate, how long the load test will be conducted, where will it be conducted, the location that the virtual users will generate from, etc… That is basic information. However, additional information will be required to smoothen the load test project and below are some of the additional considerations, you as a performance tester/test manager/consultant should be asking yourself or client. Read the rest of this entry »
Posted: April 21st, 2008 | Author: TnT Admin | Filed under: Analyze | Tags: Bottleneck, hardware | No Comments »
In this article, we will discuss two counters proposed by Mitch Tulloch to detect if the hardware is functioning properly. The two counters, System\Context Switches/sec and Processor(_Total)\% Privileged Time, both related to processor are great counters to use. In general, by using the two counters, what we want to determine the type of threads that the processor is currently working (or had been working on) and if it is busy handling interrupts instead of performing useful work. Read the rest of this entry »
Posted: April 21st, 2008 | Author: TnT Admin | Filed under: Analyze | Tags: Bottleneck, hardware | No Comments »
Back to the counters Processor(_Total)\Interrupts/sec and System\Context Switches/sec. If Processor(_Total)\Interrupts/sec does not correlate well with System\Context Switches/sec yet you noticed a sudden jump in context switches may instead suggest that your application is hitting its scalability limit on your particular machine and you may need to scale out your application (for example by clustering) or possibly redesign how it handles user mode requests. (Source: Mitch Tulloch, Windows Networking) Read the rest of this entry »
Seldom, we discuss about the planning and implementation of the load test in our site. As such, we are taking a change in this and will be touching on some considerations you as a performance tester/test manager/consultant should be thinking for every load test. These considerations enable you to smooth up the process of the entire load test and reduce any load test risks associated with it. More than often, we focus on gathering the customers requirements for the load test such as user amount to generate, how long the load test will be conducted, where will it be conducted, the location that the virtual users will generate from, etc… That is basic information. However, additional information will be required to smoothen the load test project and below are some of the additional considerations, you as a performance tester/test manager/consultant should be asking yourself or client. Read the rest of this entry »