[Kst] First time kst, plotting the grinder data

thomas at tfalkenberg.com thomas at tfalkenberg.com
Mon Aug 8 15:53:28 UTC 2011


 Hi everybody,

 this is my first approach with kst. I did some plotting with simple 
 data and everything worked fine. But I don't know how to accomplish the 
 following, more complex, use case. I have data from a load-test tool 
 (The Grinder) which has the following ASCII format:

 Thread, Run, Test, Start time (ms since Epoch), Test time, Errors, HTTP 
 response code, HTTP response length, HTTP response errors, Time to 
 resolve host, Time to establish connection, Time to first byte
 0, 0, 1013101, 1311843008835, 325, 0, 200, 5810, 0, 2, 9, 292
 0, 0, 1013100, 1311843008828, 333, 0, 0, 0, 0, 0, 0, 0
 0, 0, 1013201, 1311843012934, 103, 0, 200, 2636, 0, 2, 9, 100
 0, 0, 1013200, 1311843012933, 104, 1, 0, 0, 0, 0, 0, 0
 0, 0, 1013301, 1311843013098, 97, 0, 200, 2605, 0, 2, 9, 97
 0, 0, 1013300, 1311843013096, 99, 0, 0, 0, 0, 0, 0, 0
 0, 0, 1013401, 1311843013232, 67, 1, 200, 676, 0, 2, 9, 67
 0, 0, 1013400, 1311843013231, 68, 0, 0, 0, 0, 0, 0, 0
 0, 0, 1008101, 1311843013308, 218, 0, 200, 2044, 0, 2, 9, 218

 With the Data Wizard I can easily plot each column, using the "Start 
 Time" for the x-axis vector.
 But what I would like to have (as a first step) is a curve/image for 
 each "Test" (or even better: a configurable list of tests). E.g. a chart 
 for Test 1013101 with "Test Time" as y-value, start time as x-value. 
 Going one step further I would like to add the other colums to the 
 "Test"-charts, mainly "Errors" (preferably accumulated in 
 minute-intervals since it's only a binary value(0/1)) and "HTTP response 
 length".

 If I could automate the generation and get a live view this would 
 really be awesome, but I'm not sure if that is possible? I did think 
 about pre-processing the file and splitting it up (one for each "Test"), 
 but I would like to avoid that if possible.

 Is this a good scenario for kst or should I maybe look at other tools?

 Thanks a lot for your help,
 Thomas



More information about the Kst mailing list