Hi There Stéphane,
I’m currently using Gatling splitting load injection using multiple pods in a k8s infra, but I’m facing intermittent issues. Basically for some tests load injectors are consuming a lot of cpu and k8s kills them. I have no cpu limits configured for these load injectors and can’t understand what can causing the issue.
Load test profile is the same during the test.
Response payloads have ~2MB and I’m performing 2 jsonpath over the response as follows:
Target.buildGETRequest("/")
.check(
jsonPath("$.element[?(@.type == ‘TYPE_A’)].links[?(@.rel == ‘next’)].href").findRandom.optional.saveAs(“hrefA”),
jsonPath("$.channels[?(@.type == ‘TYPE_B’)].links[?(@.rel == ‘next’)].href").findRandom.optional.saveAs(“hrefB”)
)
Could this be the cause for this scenario?
There is some debug I can perform in load injector side to troubleshoot this?
Thanks for your help,
Nuno Marcos
