id | title | sidebar_label |
---|---|---|
disk-fill |
Disk Fill Experiment Details |
Disk Fill |
Type | Description | Tested K8s Platform |
---|---|---|
Chaos | Fill up Ephemeral Storage of a Pod | GKE, EKS, AKS |
- Ensure that Kubernetes Version > 1.16
- 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
disk-fill
experiment resource is available in the cluster by executingkubectl get chaosexperiments
in the desired namespace If not, install from here - Cluster must run docker container runtime
- Either an appropriate Ephemeral Storage Requests and Limits should be set for the application or
EPHEMERAL_STORAGE_MEBIBYTES
ENV should be defined in the ChaosEngine before running the experiment. An example specification is shown below:
apiVersion: v1
kind: Pod
metadata:
name: frontend
spec:
containers:
- name: db
image: mysql
env:
- name: MYSQL_ROOT_PASSWORD
value: "password"
resources:
requests:
ephemeral-storage: "2Gi"
limits:
ephemeral-storage: "4Gi"
- name: wp
image: wordpress
resources:
requests:
ephemeral-storage: "2Gi"
limits:
ephemeral-storage: "4Gi"
- Application pods are healthy before chaos injection.
- Application pods are healthy post chaos injection.
- Causes Disk Stress by filling up the ephemeral storage of the pod on any given node.
- Causes the application pod to get evicted if the capacity filled exceeds the pod's ephemeral storage limit.
- Tests the Ephemeral Storage Limits, to ensure those parameters are sufficient.
- Tests the application's resiliency to disk stress/replica evictions.
- Disk Fill can be effected using the chaos library:
litmus
, which makes use ofdd
to create a file of specified capacity on the node. - The desired chaoslib can be selected by setting the above options as value for the env variable
LIB
-
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.
- 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.
---
apiVersion: v1
kind: ServiceAccount
metadata:
name: disk-fill-sa
namespace: default
labels:
name: disk-fill-sa
app.kubernetes.io/part-of: litmus
---
apiVersion: rbac.authorization.k8s.io/v1
kind: Role
metadata:
name: disk-fill-sa
namespace: default
labels:
name: disk-fill-sa
app.kubernetes.io/part-of: litmus
rules:
- apiGroups: [""]
resources: ["pods","events"]
verbs: ["create","list","get","patch","update","delete","deletecollection"]
- apiGroups: [""]
resources: ["pods/exec","pods/log","replicationcontrollers"]
verbs: ["list","get","create"]
- apiGroups: ["batch"]
resources: ["jobs"]
verbs: ["create","list","get","delete","deletecollection"]
- apiGroups: ["apps"]
resources: ["deployments","statefulsets","daemonsets","replicasets"]
verbs: ["list","get"]
- apiGroups: ["apps.openshift.io"]
resources: ["deploymentconfigs"]
verbs: ["list","get"]
- apiGroups: ["argoproj.io"]
resources: ["rollouts"]
verbs: ["list","get"]
- apiGroups: ["litmuschaos.io"]
resources: ["chaosengines","chaosexperiments","chaosresults"]
verbs: ["create","list","get","patch","update"]
---
apiVersion: rbac.authorization.k8s.io/v1
kind: RoleBinding
metadata:
name: disk-fill-sa
namespace: default
labels:
name: disk-fill-sa
app.kubernetes.io/part-of: litmus
roleRef:
apiGroup: rbac.authorization.k8s.io
kind: Role
name: disk-fill-sa
subjects:
- kind: ServiceAccount
name: disk-fill-sa
namespace: default
Note: In case of restricted systems/setup, create a PodSecurityPolicy(psp) with the required permissions. The chaosServiceAccount
can subscribe to work around the respective limitations. An example of a standard psp that can be used for litmus chaos experiments can be found here.
- Provide the application info in
spec.appinfo
- Provide the auxiliary applications info (ns & labels) in
spec.auxiliaryAppInfo
- Override the experiment tunables if desired in
experiments.spec.components.env
- To understand the values to provided in a ChaosEngine specification, refer ChaosEngine Concepts
Variables | Description | Specify In ChaosEngine | Notes |
---|---|---|---|
FILL_PERCENTAGE | Percentage to fill the Ephemeral storage limit | Mandatory | Can be set to more than 100 also, to force evict the pod |
TARGET_CONTAINER | Name of container which is subjected to disk-fill | Optional | If not provided, the first container in the targeted pod will be subject to chaos |
CONTAINER_PATH | Storage Location of containers | Optional | Defaults to '/var/lib/docker/containers' |
TOTAL_CHAOS_DURATION | The time duration for chaos insertion (sec) | Optional | Defaults to 60s |
TARGET_PODS | Comma separated list of application pod name subjected to disk fill chaos | Optional | If not provided, it will select target pods randomly based on provided appLabels |
DATA_BLOCK_SIZE | It contains data block size used to fill the disk(in KB) | Optional | Defaults to 256, it supports unit as KB only |
PODS_AFFECTED_PERC | The Percentage of total pods to target | Optional | Defaults to 0 (corresponds to 1 replica), provide numeric value only |
LIB | The chaos lib used to inject the chaos | Optional | Defaults to `litmus` supported litmus only |
LIB_IMAGE | The image used to fill the disk | Optional | Defaults to `litmuschaos/go-runner:latest` |
RAMP_TIME | Period to wait before injection of chaos in sec | Optional | |
SEQUENCE | It defines sequence of chaos execution for multiple target pods | Optional | Default value: parallel. Supported: serial, parallel |
EPHEMERAL_STORAGE_MEBIBYTES | Ephemeral storage which need to fill (unit: MiBi) | Optional | |
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 |
apiVersion: litmuschaos.io/v1alpha1
kind: ChaosEngine
metadata:
name: nginx-chaos
namespace: default
spec:
# 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: disk-fill-sa
experiments:
- name: disk-fill
spec:
components:
env:
- name: TOTAL_CHAOS_DURATION
value: '60'
# specify the fill percentage according to the disk pressure required
- name: FILL_PERCENTAGE
value: '80'
- name: PODS_AFFECTED_PERC
value: ''
# Provide the container runtime path
# Default set to docker container path
- name: CONTAINER_PATH
value: '/var/lib/docker/containers'
-
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.
-
View the status of the pods as they are subjected to disk stress.
watch -n 1 kubectl get pods -n <application-namespace>
-
Monitor the capacity filled up on the host filesystem
watch -n 1 du -kh /var/lib/docker/containers/<container-id>
-
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"}}'
Notes:
-
The abort will stop further fill of the local disk, but doesn't reclaim the used capacity. This is a manual operation as of today. The auto-rollback, i.e., in this case reclaim of currently filled disk-space will be implemented in a future release
-
However, upon graceful completion of the experiment (i.e.,un-aborted), the space is automatically reclaimed as the chaos impact is reverted.
-
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-disk-fill -n <application-namespace>
- A sample recording of this experiment execution is provided here.