Container Kill Experiment Details
Experiment Metadata
Type | Description | Tested K8s Platform |
---|---|---|
Generic | Kill one container in the application pod | GKE, Packet(Kubeadm), Minikube, EKS, AKS |
Prerequisites
- Ensure that the Litmus Chaos Operator is running by executing
kubectl get pods
in operator namespace (typically,litmus
). If not, install from here - Ensure that the
container-kill
experiment resource is available in the cluster by executingkubectl get chaosexperiments
in the desired namespace. If not, install from here
Entry Criteria
- Application pods are healthy before chaos injection
Exit Criteria
- Application pods are healthy post chaos injection
Details
- litmus lib in docker runtime details
- It can kill the container of multiple pods in parellel (can be tuned by
PODS_AFFECTED_PERC
env). It kill the container by sending SIGKILL termination signal to its docker socket (hence docker runtime is required) - Containers are killed using the
kill
command provided by pumba - Pumba is run as a pod on the application node. It have ability to kill the application containers multiple times. Which can be varied by
TOTAL_CHAOS_DURATION
andCHAOS_INTERVAL
.
- It can kill the container of multiple pods in parellel (can be tuned by
- litmus chaoslib in containerd and crio runtime codetails
- It can kill the container of multiple pods in parellel (can be tuned by
PODS_AFFECTED_PERC
env). - Containers are killed using the
crictl stop
command. - container-kill is run as a pod on the application node. It have ability to kill the application containers multiple times. Which can be varied by
TOTAL_CHAOS_DURATION
andCHAOS_INTERVAL
.
- It can kill the container of multiple pods in parellel (can be tuned by
- Tests deployment sanity (replica availability & uninterrupted service) and recovery workflow of the application
- Good for testing recovery of pods having side-car containers
Integrations
- Container kill is achieved using the
litmus
chaos library - The container runtime can be choose via setting
CONTAINER_RUNTIME
env. supported values:docker
,containerd
,crio
- The desired pumba and litmus image can be configured in the env variable
LIB_IMAGE
.
Steps to Execute the Chaos Experiment
This Chaos Experiment can be triggered by creating a ChaosEngine resource on the cluster. To understand the values to provide in a ChaosEngine specification, refer Getting Started
Follow the steps in the sections below to create the chaosServiceAccount, prepare the ChaosEngine & execute the experiment.
Prepare chaosServiceAccount
- Use this sample RBAC manifest to create a chaosServiceAccount in the desired (app) namespace. This example consists of the minimum necessary role permissions to execute the experiment.
Sample Rbac Manifest
---
apiVersion: v1
kind: ServiceAccount
metadata:
name: container-kill-sa
namespace: default
labels:
name: container-kill-sa
---
apiVersion: rbac.authorization.k8s.io/v1beta1
kind: Role
metadata:
name: container-kill-sa
namespace: default
labels:
name: container-kill-sa
rules:
- apiGroups: ["","litmuschaos.io","batch","apps"]
resources: ["pods","jobs","pods/exec","pods/log","events","chaosengines","chaosexperiments","chaosresults"]
verbs: ["create","list","get","patch","update","delete","deletecollection"]
---
apiVersion: rbac.authorization.k8s.io/v1beta1
kind: RoleBinding
metadata:
name: container-kill-sa
namespace: default
labels:
name: container-kill-sa
roleRef:
apiGroup: rbac.authorization.k8s.io
kind: Role
name: container-kill-sa
subjects:
- kind: ServiceAccount
name: container-kill-sa
namespace: default
Prepare ChaosEngine
- Provide the application info in
spec.appinfo
- Override the experiment tunables if desired in
experiments.spec.components.env
- To understand the values to provided in a ChaosEngine specification, refer ChaosEngine Concepts
Supported Experiment Tunables
Variables | Description | Specify In ChaosEngine | Notes | |
---|---|---|---|---|
TARGET_CONTAINER | The container to be killed inside the pod | Mandatory | If the TARGET_CONTAINER is not provided it will delete the first container | |
CHAOS_INTERVAL | Time interval b/w two successive container kill (in sec) | Optional | If the CHAOS_INTERVAL is not provided it will take the default value of 10s | |
TOTAL_CHAOS_DURATION | The time duration for chaos injection (seconds) | Optional | Defaults to 20s | |
PODS_AFFECTED_PERC | The Percentage of total pods to target | Optional | Defaults to 0% (corresponds to 1 replica) | |
TARGET_POD | Name of the application pod subjected to container kill chaos | Optional | If not provided it will select from the app label provided | |
LIB_IMAGE | LIB Image used to kill the container | Optional | Defaults to `litmuschaos/go-runner:latest` | |
LIB | The category of lib use to inject chaos | Optional | Default value: litmus, only litmus supported | |
RAMP_TIME | Period to wait before injection of chaos in sec | Optional | ||
SOCKET_PATH | Path of the containerd/crio socket file | Optional | Defaults to `/run/containerd/containerd.sock` | |
CONTAINER_RUNTIME | container runtime interface for the cluster | Optional | Defaults to docker, supported values: docker, containerd, crio | |
INSTANCE_ID | A user-defined string that holds metadata/info about current run/instance of chaos. Ex: 04-05-2020-9-00. This string is appended as suffix in the chaosresult CR name. | Optional | Ensure that the overall length of the chaosresult CR is still < 64 characters |
Sample ChaosEngine Manifest
apiVersion: litmuschaos.io/v1alpha1
kind: ChaosEngine
metadata:
name: nginx-chaos
namespace: default
spec:
# It can be true/false
annotationCheck: 'true'
# It can be active/stop
engineState: 'active'
#ex. values: ns1:name=percona,ns2:run=nginx
auxiliaryAppInfo: ''
appinfo:
appns: 'default'
applabel: 'app=nginx'
appkind: 'deployment'
chaosServiceAccount: container-kill-sa
monitoring: false
# It can be delete/retain
jobCleanUpPolicy: 'delete'
experiments:
- name: container-kill
spec:
components:
env:
# specify the name of the container to be killed
- name: TARGET_CONTAINER
value: 'nginx'
# provide the chaos interval
- name: CHAOS_INTERVAL
value: '10'
# provide the total chaos duration
- name: TOTAL_CHAOS_DURATION
value: '20'
# provide the name of container runtime
# it supports docker, containerd, crio
# default to docker
- name: CONTAINER_RUNTIME
value: 'docker'
# provide the socket file path
# applicable only for containerd runtime
- name: SOCKET_PATH
value: '/run/containerd/containerd.sock'
Create the ChaosEngine Resource
Create the ChaosEngine manifest prepared in the previous step to trigger the Chaos.
kubectl apply -f chaosengine.yml
If the chaos experiment is not executed, refer to the troubleshooting section to identify the root cause and fix the issues.
Watch Chaos progress
View pod restart count by setting up a watch on the pods in the application namespace
watch -n 1 kubectl get pods -n <application-namespace>
Abort/Restart the Chaos Experiment
To stop the pod-delete experiment immediately, either delete the ChaosEngine resource or execute the following command:
kubectl patch chaosengine <chaosengine-name> -n <namespace> --type merge --patch '{"spec":{"engineState":"stop"}}'
To restart the experiment, either re-apply the ChaosEngine YAML or execute the following command:
kubectl patch chaosengine <chaosengine-name> -n <namespace> --type merge --patch '{"spec":{"engineState":"active"}}'
Check Chaos Experiment Result
Check whether the application is resilient to the container kill, once the experiment (job) is completed. The ChaosResult resource name is derived like this:
<ChaosEngine-Name>-<ChaosExperiment-Name>
.kubectl describe chaosresult nginx-chaos-container-kill -n <application-namespace>
Application Container Kill Demo
- A sample recording of this experiment execution is provided here.