zooz / predator-runner Goto Github PK

As a user i'd like to create resources random data like email, names and etc.

This can be simply done by adding chance.js to the runner.

I0225 15:09:53.460655 26258 exec.cpp:162] Version: 1.2.3
I0225 15:09:53.464668 26272 exec.cpp:237] Executor registered on agent 8fa373fb-cbf2-4fce-8051-53d4077cdde5-S32
/usr/app/utils/verifier.js:32
return mandatoryVariables.filter((variable) => {
^

TypeError: Cannot read property ‘filter’ of undefined
at getMissingVariables (/usr/app/utils/verifier.js:32:31)
at Object.module.exports.verifyEnvironmentVars (/usr/app/utils/verifier.js:20:35)
at Object. (/usr/app/app.js:3:29)
at Module._compile (module.js:653:30)
at Object.Module._extensions..js (module.js:664:10)
at Module.load (module.js:566:32)
at tryModuleLoad (module.js:506:12)
at Function.Module._load (module.js:498:3)
at Function.Module.runMain (module.js:694:10)
at startup (bootstrap_node.js:204:16)

failing with PROMETHEUS and not prometheus

Hi, yesterday I upgraded from 1.5 to 1.6.0. I found status of pod of runner always shows Error status like following

predator.a3733d0d-5a00-443d-a956-f99be870c5fa-9tpns               0/1     Error 

the logs showed

{"name":"predator-runner","hostname":"predator.a3733d0d-5a00-443d-a956-f99be870c5fa-9tpns","pid":1,"level":30,"msg":"Runner exceeded test duration, sending DONE status and existing","time":"2021-02-17T04:26:48.467Z","v":0}

Besides, I can saw the results of the test during the test but after the test finished. The test status turned to Failed so that I cannot see the result anymore.

I noticed that in the Dockerfile, this is present: --max_old_space_size=192

This causes node.js to crash (out of heap memory) when the processor requires memory to load the necessary "data" for the tests.

I propose that we make this value dynamic so it can be customised.

I have a processor which performs a query at the beginning of a scenario and loads a number of user tokens (JWT) and other records needed for the load test, and with this limit the job gets interrupted when one of the pod reaches that limit. It stopped doing that when I changed 192 to 512.

What do you think?

Describe the bug
In some configurations the runner keeps sending requests after the test duration is up, in this case the runner should send a DONE stat and the test should be marked as finished

To Reproduce
Steps to reproduce the behavior:

1. Run 2000 RPS test with 1 parallism on a very fast API server(5ms response rate)
2. No avoided scenarios and therefore the runner will keep sending requests after the time is up

Expected behavior

1. Test finish in the expected time duration

Version:

• Predator: 1.6.0
• Predator-runner: 1.6.0

Additional context
Possibility to add this implementation with an environment variable in the predator installation, some users might not want the test to finish and let it keep running