I am working on Apache Jmeter. I am going to simulate DOS attack with this software, and measure the Response Time / Experiment Time. I should produce these factors with Graph. How should I do it?
Read this:
http://www.ubik-ingenierie.com/blog/automatically-generating-nice-graphs-at-end-of-your-load-test-with-apache-jmeter-and-jmeter-plugins/
It uses JMeter Plugins that you easily install following this procedure.
Related
How I analyze jmeter reports. Where is the best tutorial available
If you already executed your test and generated HTML Reporting Dashboard you should see some numbers in the statistics table.
These are so called KPIs, see JMeter Glossary for the description on what do the metrics mean.
The next step would be detecting KPI correlations, like see the relationship between increasing number of users and the increasing response time or errors rate.
And the final step would be creating a report which normally consists of two parts:
Manager-friendly summary stating where the application meets NFR, what is the saturation point, what is the first bottleneck, what are possible risks, etc.
Technical details with in-depth analysis of performance degradations including logs, slow database queries, profiler tool output interpretation, etc.
In a network with some wireless nodes, after simulation, the Omnet++ generates a scalar file.
I want to calculate throughput, goodput, end-to-end delay time, packets lost count from the generated scalar file.
Is there a tools that calculate them? or script?
If there isn't, what is the best solution?
I search similar question in the stackoverflow site, most questions are unanswered.
Assuming you are using INET, if the application layer emits a packetReceived signal then you can use the dataAge(packetReceived) and throughput(packetReceived) result filters.
This will produce these results in the .sca and .vec files.
For an example of how to use them see: https://github.com/inet-framework/inet/blob/007bc454ec7749e2dea8fcb808429d21074880ad/src/inet/applications/udpapp/UdpSink.ned
Some example applications which will work with this are: IpvxTrafGen, SctpClient/Peer, TcpAppBase (and derivatives), UdpBasicApp/Burst, UdpSink etc.
I am using the JMeter tool to evaluate the performance testing of several SPARQL queries. I use the "Sampler" and "Http request" section of Jmeter (like this image ) to write the SPARQL queries. However it does not work as JMeter does not give me the actual response time of the queries. The more complex query has the same response time or even less. Maybe I have to use other options of JMeter.
Does anybody have an experience on this?
You may be interested in the SPARQL Query Benchmarker tool (disclaimer - I am the primary author on that tool) which is specifically designed for this purpose
It distinguishes between response time as being the time when the first result is received from the remote system and runtime as the time taken to receive all results. The Benchmarking Methodology page describes more about the difference between these metrics and the various other metrics which the tool calculates.
Since many systems execute queries and generate results in a streaming fashion for complex queries you will typically see noticeable differences between these two figures. Of course systems vary so not all systems will display quite the same variations between the two metrics.
I have measured separately our application server CPU with 5 users and 15 users. Both runs took 5 minutes. I would like to show both runs and CPU in one graph. Currently the only way to do it is to export CSV to Excel and create custom graphs. This is silly because the graphs are quite good in JMeter itself.
JMeter plugin extensions contains a graph called Composite graph, but the use-case seems to be to be able to add many data points to one and the same graph from within the same run. It doesn't seem to be able to merge graphs from two separate thread runs because the time in the two are different (the two runs have been executed in succession and therefore the graphs are shown after one-another, not at the same time).
Any ideas? Thanks!
It looks like it can be done using the information here jmeter_wiki but that it will still require some effort.
i have a combinatoric problem as such:
You are given N testers.
Each tester is one of M different types.
Each tester can be configured to use one of P different configs.
.
You have L lots of products to test,
Each product can only be tested on specific Tester type,
Each product can only be tested by Tester configured with specific Configs. Some of the Configs can be applied on multiple products.
Any tester can change its config during production, but each change on tester config will incur additional time U .
Each lot has a lot size that determines its test-time, Q.
Now i need to come out a lot scheduling algorithm such that the time to finish testing all lots is minimum.
What are the best approaches to tackle this kind of problem ?
It can be modelled as a Job-Shop problem (JSP) with setup times where the job size is 1. Unfortunately it gets pretty hard to find an optimum when the number of jobs > 10.
There are many free solver implementations out there that contain Job-Shop as a sample problem:
If you're using C++, Gecode is good. If you're free to choose, ECLiPSe Prolog contains source code for the JSP.
If you can do with a good solution (instead of an optimal one), I suggest using a greedy algorithm (for the JSP, greedy algorithms give solutions typically within 10% of the optimum - I had some experience with this). I'm gonna think about one and get back here (the problem is what are called 'setup time constraints', i.e. the constraints coming from changing tester configuration).