kenesparta / fullcycle-stress-test

Stress Test CLI

Stress test CLI

🚀 Starting the application!

• Execute the command make run. and then we can see the container set in docker-compose.yaml.

🧪 How can I test?

• You need to execute this command

  docker compose run --rm stress-test-cli -h

• Then you can see this output:

  Usage of ./stress-test-cli:
  -m string
      HTTP method to use (default "GET")
  -r int
      Amount of requests to send (default 10)
  -url string
      URL from service to test
  -w int
      Amount of concurrent requests (workers) (default 1)
  

• After that, you can execute this command (you should install jq)

  docker compose run --rm stress-test-cli -url http://google.com/ | jq

• You can use your custom flags

    docker compose run --rm stress-test-cli -url http://google.com/ -w 20 -r 1000 | jq | jq

• After you execute the command, you can se a response, something like this:

  {
   "requests": 1000,
   "workers": 20,
   "status_map_count": {
     "200": 701,
     "301": 3,
     "302": 79,
     "429": 217
   },
   "errors": [],
   "error_count": 0,
   "total_duration": "33.274527583s"
  }

  • requests: the total amount of requests performed. This number must be equal to the sum of all the status_map_count values and error_count value.
  • workers: the total amount of concurrency requests performed.
  • status_map_count: the result map that counts the amount of HTTP requests status code, in this example we have 701 requests with status 200, 3 requests with the status 301, 79 with 302 and 217 with 409.
  • errors: The array of errors on the request. These errors are related with each request.
  • error_count: the amount of errors that we have during the stress tests.
  • total_duration: the duration in string format, for this case, we perform the test in ~33s.

🚨 Errors

We have identified these possible errors that validated the input params BEFORE we proceed with the execution of the stress tests.

Condition	Example	Json Response
r <= 0	-w 12 -r 0	{"error": "request should be positive number"}
w <= 0	-w 0 -r 13	{"error": "concurrency value should be positive number"}
w > r	-w 500 -r 5	{"error": "we have more workers (go routines) than requests"}
valid(url)	-url hhpp:/g	{"error": "should be a URL valid value"}
validMethod(m)	-m PMTD	{"error": "should be a valid HTTP method"}

Type of errors

Error before executing the request tests

If we execute this:

docker compose run --rm stress-test-cli -url http://google.com/ -w 500 -r 5 | jq

we show this error:

{
  "error": "we have more workers (go routines) than requests"
}

Since w > r, as indicated in the previously shown table, we are in a blocked condition, and therefore, we will not proceed with the execution of the stress test.

Errors in each request

If we execute this:

docker compose run --rm stress-test-cli -url http://google.com/ -w 5 -r 500 | jq

we have

{
  "requests": 500,
  "workers": 5,
  "status_map_count": {
    "200": 498
  },
  "errors": [
    "Get \"http://www.google.com/\": stopped after 10 redirects",
    "Get \"http://www.google.com/\": stopped after 10 redirects"
  ],
  "error_count": 2,
  "total_duration": "4.470558169s"
}

the errors in the JSON are showing because we have these errors in each request.