mirrors/kyverno
1
0
Fork 0
mirror of https://github.com/kyverno/kyverno.git synced 2024-12-14 11:57:48 +00:00
Code Activity
kyverno/litmuschaos
History
shuting d5f6167e56
Fix flaky e2e tests for generate policies (#3681) 
* fix flaky generate e2e tests

Signed-off-by: ShutingZhao <shuting@nirmata.com>

* enable validate, verifyimage e2e tests

Signed-off-by: ShutingZhao <shuting@nirmata.com>

* set policy names different within a single test

Signed-off-by: ShutingZhao <shuting@nirmata.com>

* do not delete cloned resource when sync generate policy is deleted

Signed-off-by: ShutingZhao <shuting@nirmata.com>

* replace grLister by urLister

Signed-off-by: ShutingZhao <shuting@nirmata.com>

* re-queue pending URs only to fix clone policy deletion

Signed-off-by: ShutingZhao <shuting@nirmata.com>

* remove duplicate import

Signed-off-by: ShutingZhao <shuting@nirmata.com>

Co-authored-by: Sambhav Kothari <sambhavs.email@gmail.com>
Co-authored-by: Charles-Edouard Brétéché <charled.breteche@gmail.com>
2022-04-26 19:18:24 +00:00
..
pod_cpu_hog Fix flaky e2e tests for generate policies (#3681) 2022-04-26 19:18:24 +00:00
README.md Integrate LitmusChaos - Pod Memory Hog experiment (#2014) 2021-06-23 14:16:49 -07:00

README.md

Integration of Kyverno with Litmus

Kyverno is a policy engine designed for Kubernetes. It can validate, mutate, and generate configurations using admission controls and background scans. Litmus provides a large number of experiments for testing containers, pods, nodes, as well as specific platforms and tools. The advantage of chaos engineering is that one can quickly figure out issues that other testing layers cannot easily capture. This can save a lot of time in the future, and will help to find the loopholes in the system and fix them.

Steps to Execute LitmusChaos Experiment

Prerequisites

  • Ensure that Kubernetes Version > 1.15
  • Ensure that the Kyverno is running by executing kubectl get pods in operator namespace (typically, kyverno). If not, install from here.
  • Update Kyverno Deployment to use ghcr.io/kyverno/kyverno:test-litmuschaos image. Note that this image is built specifically to run Litmuschaos experiments per this request, CHAOS_KILL_COMMAND. The official Kyverno images will adopt this soon.
  • Ensure that the Litmus Chaos Operator is running by executing kubectl get pods in operator namespace (typically, litmus). If not, install from here.

Running experiment

Aftr setting up the docker images, for running a LitmusChaos experiment following steps need to be followed -

  • Firstly, exicute eval export E2E="ok"
  • Run the Chaos Experiment Test Command - go test ./litmuschaos/pod_cpu_hog -v.

The test passes if the enforce policy shows it's expected behaviour.

Experiments

Experiment name LitmusChaos experiment - Pod CPU Hog
Test command go test ./litmuschaos/pod_cpu_hog -v
Goal Seeing how the overall application stack behaves when Kyverno pods experience CPU spikes either due to expected/undesired processes
Performed tests
  • Deploy enforce policy.
  • Run the chaos test to consume CPU resources on the application container.
  • Verify the enforce policy behaviour.
    • Expected result Kyverno pod is responding after running Pod CPU Hog Experiment