HTTP Benchmark Example

Step 1—new benchmark

Run the following command to create the Java version of the example benchmark program:

  • jbang tulip-cli@wfouche init Java

This command creates various files and directories:

├── benchmark_config.json
├── io
│   └── tulip
│       ├── App.java
│       └── HttpUser.java
├── run_bench.cmd
└── run_bench.sh

Step 2—run benchmark

Run script run_bench.sh (Linux or macOS), or run_bench.cmd (Windows) to start the benchmark program.

While the benchmark program is running the benchmark_output.json file is incrementally updated with benchmark results.

Step 3—view benchmark results

├── benchmark_config.adoc
├── benchmark_config.html
├── benchmark_config.json
├── benchmark_output.json
├── benchmark_report.html

── benchmark_report.json ├── io │   └── tulip │   ├── App.java │   └── HttpUser.java ├── run_bench.cmd └── run_bench.sh














Once the benchmark is finished, a benchmark report is automatically generated from the benchmark_output.json file.
The benchmark report file is called benchmark_report.html. The benchmark script displays the benchmark report on the console using the lynx text web browser utility (as shown below).















image demo benchmark report console












Open the benchmark_report.html file in a browser to view the benchmark results.















image demo benchmark report html
















Step 4—view configuration






The configuration report is stored in file benchmark_config.html.















image demo benchmark config report 1
















Step 5—view JSON summary






File benchmark_report.json contains the same results that are displayed in the HTML benchmark report. Having the results in a JSON file format, makes it easy to validate the final results.






json_report.py




import json
from collections import OrderedDict

filename = "benchmark_report.json"
fileObj = open(filename)
jb = json.load(fileObj, object_pairs_hook=OrderedDict)

def report(name):
  print name
  print "  ", jb["benchmarks"][name]["actions"]["summary"]["aps"], "aps"
  print "  ", jb["benchmarks"][name]["actions"]["summary"]["aps_target_rate"], "aps_target_rate"

report("REST1")
report("REST2")
report("REST3")
report("REST3.max")








$ jbang run jython-cli@jython json_report.py








REST1
   10.0 aps
   10.0 aps_target_rate
REST2
   10.0 aps
   10.0 aps_target_rate
REST3
   10.0 aps
   10.0 aps_target_rate
REST3.max
   117.977777778 aps
   0.0 aps_target_rate