- Kubernetes cheatsheet
- Getting Started
- Sample yaml
- Workflow
- Physical components
- Everything is an object - persistent entities
- Notes
- Labs
- Fault tolerance
- Rollback
- Auto-healing
- Auto-scaling
- Load-balancing
- Isolation (sandbox)
apiVersion: <>
kind: <>
metadata:
name: <>
labels:
...
annotations:
...
spec:
containers:
...
initContainers:
...
priorityClassName: <>Credit: https://www.reddit.com/user/__brennerm/
-
(kube-scheduler, controller-manager, etcd) --443--> API Server
-
API Server --10055--> kubelet
- non-verified certificate
- MITM
- Solution:
- set kubelet-certificate-authority
- ssh tunneling
-
API server --> (nodes, pods, services)
- Plain HTTP (unsafe)
- API Server (443)
- kube-scheduler
- controller-manager
- cloud-controller-manager
- kube-controller-manager
- etcd
Other components talk to API server, no direct communication
-
Kubelet
-
Container Engine
- CRI
- The protocol which used to connect between Kubelet & container engine
- CRI
-
Kube-proxy
-
maintained in etcd, identified using
- names: client-given
- UIDs: system-generated
-
Both need to be unique
-
three management methods
- Imperative commands (kubectl)
- Imperative object configuration (kubectl + yaml)
- repeatable
- observable
- auditable
- Declarative object configuration (yaml + config files)
- Live object configuration
- Current object configuration file
- Last-applied object configuration file
Node Capacity
---------------------------
| kube-reserved |
|---------------------------|
| system-reserved |
| ------------------------- |
| eviction-threshold |
| ------------------------- |
| |
| allocatable |
| (available for pods) |
| |
| |
---------------------------
-
Three pre-defined
- default
- kube-system
- kube-public: auto-readable by all users
-
Objects without namespaces
- Nodes
- PersistentVolumes
- Namespaces
- key / value
- loose coupling via selectors
- need not be unique
- Independent of lifespan of any backend pod
- Service object has a static port assigned to it
- ReplicaSet, deployment, daemonset, statefulSet
- Actual state <-> desired state
- reconciliation loop
- nodeSelector
- Affinity & Anti-Affinity
- Node
- Steer pod to node
- Pod
- Steer pod towards or away from pods
- Node
- Taints & tolerations (anti-affinity between node and pod!)
- Base on predefined configuration (env=dev:NoSchedule)
... tolerations: - key: "dev" operator: "equal" value: "env" effect: NoSchedule ...
- Base on node condition (alpha in v1.8)
- taints added by node controller
- Base on predefined configuration (env=dev:NoSchedule)
kubectl run name --image=<image>What's available inside the container?
- File system
- Image
- Associated Volumes
- ordinary
- persistent
- Container
- Hostname
- Pod
- Pod name
- User-defined envs
- Services
- List of all services
Access with:
-
Symlink (important):
- /etc/podinfo/labels
- /etc/podinfo/annotations
-
Or:
volumes:
- name: podinfo
downwardAPI:
items:
- path: "labels"
fieldRef:
fieldPath: metadata.labels
- path: "annotations"
fieldRef:
fieldPath: metadata.annotations- Pending
- Running
- Succeeded
- Failed
- Unknown
- Liveness
- Failed? Restart policy applied
- Readiness
- Failed? Removed from service
- available since 1.8
- PriorityClass object
- Affect scheduling order
- High priority pods could jump the queue
- Preemption
- Low priority pods could be pre-empted to make way for higher one (if no node is available for high priority)
- These preempted pods would have a graceful termination period
- Share access to memory space
- Connect to each other using localhost
- Share access to the same volume
- entire pod is host on the same node
- all in or nothing
- no auto healing or scaling
- run before app containers
- always run to completion
- run serially
- PostStart
- PreStop (blocking)
Handlers:
- Exec
- HTTP
...
spec:
containers:
lifecycle:
postStart:
exec:
command: <>
preStop:
http:
...Could invoke multiple times
When Kubernetes creates a Pod it assigns one of these QoS classes to the Pod:
- Guaranteed (all containers have limits == requests)
If a Container specifies its own memory limit, but does not specify a memory request, Kubernetes automatically assigns a memory request that matches the limit. Similarly, if a Container specifies its own cpu limit, but does not specify a cpu request, Kubernetes automatically assigns a cpu request that matches the limit.
- Burstable (at least 1 has limits or requests)
- BestEffort (no limits or requests)
You can use a podpreset object to inject information like secrets, volume mounts, and environment variables etc into pods at creation time. This task shows some examples on using the PodPreset resource
apiVersion: settings.k8s.io/v1alpha1
kind: PodPreset
metadata:
name: allow-database
spec:
selector:
matchLabels:
role: frontend
env:
- name: DB_PORT
value: "6379"
volumeMounts:
- mountPath: /cache
name: cache-volume
volumes:
- name: cache-volume
emptyDir: {}Features:
- Scaling and healing
- Pod template
- number of replicas
Components:
-
Pod template
-
Pod selector (could use matchExpressions)
-
Label of replicaSet
-
Number of replica
-
Could delete replicaSet without its pods using
--cascade =false -
Isolating pods from replicaSet by changing its labels
-
versioning and rollback
-
Contains spec of replicaSet within it
-
advanced deployment
-
blue-green
-
canary
-
Update containers --> new replicaSet & new pods created --> old RS still exists --> reduced to zero
-
Every change is tracked
-
Append
--recordin kubectl to keep history -
Update strategy
- Recreate
- Old pods would be killed before new pods come up
- RollingUpdate
- progressDeadlineSeconds
- minReadySeconds
- rollbackTo
- revisionHistoryLimit
- paused
- spec.Paused
- Recreate
-
kubectl rollout undo deployment/<> --to-revision=<> -
kubectl rollout statua deployment/<> -
kubectl set image deployment/<> <>=<>:<> -
kubectl rollout resume/pause <>
- RC = ( RS + deployment ) before
- Obsolete
- Ensure all nodes run a copy of pod
- Cluster storage, log collection, node monitor ...
- Maintains a sticky identity
- Not interchangeable
- Identifier maintains across any rescheduling
Limitation
- volumes must be pre-provisioned
- Deleting / Scaling will not delete associated volumes
Flow
- Deployed 0 --> (n-1)
- Deleted (n-1) --> 0 (successor must be completely shutdown before proceed)
- Must be all ready and running before scaling happens
- Non-parallel jobs
- Parallel jobs
- Fixed completion count
- job completes when number of completions reaches target
- With work queue
- requires coordination
- Fixed completion count
- Use spec.activeDeadlineSeconds to prevent infinite loop
- Job should be idempotent
- Targets: replicaControllers, deployments, replicaSets
- CPU or custom metrics
- Won't work with non-scaling objects: daemonSets
- Prevent thrashing (upscale/downscale-delay)
Credit: https://www.reddit.com/user/__brennerm/
-
Logical set of backend pods + frontend
-
Frontend: static IP + port + dns name
-
Backend: set of backend pods (via selector)
-
Static IP and networking.
-
Kube-proxy route traffic to VIP.
-
Automatically create endpoint based on selector.
-
CluterIP
-
NodePort
- external --> NodeIP + NodePort --> kube-proxy --> ClusterIP
-
LoadBalancer
- Need to have cloud-controller-manager
- Node controller
- Route controller
- Service controller
- Volume controller
- external --> LB --> NodeIP + NodePort --> kube-proxy --> ClusterIP
- Need to have cloud-controller-manager
-
ExternalName
- Can only resolve with kube-dns
- No selector
Service discovery
- SRV record for named port
- port-name.port-protocol.service-name.namespace.svc.cluster.local
- Pod domain
- pod-ip-address.namespace.pod.cluster.local
- hostname is
metadata.name
spec.dnsPolicy
- default
- inherit node's name resolution
- ClusterFirst
- Any DNS query that does not match the configured cluster domain suffix, such as “www.kubernetes.io”, is forwarded to the upstream nameserver inherited from the node
- ClusterFirstWithHostNet
- if host network = true
- None (since k8s 1.9)
- Allow custom dns server usage
Headless service
- with selector? --> associate with pods in cluster
- without selector? --> forward to externalName
Could specify externalIP to service
Credit: https://www.reddit.com/user/__brennerm/
Lifetime longer than any containers inside a pod.
4 types:
-
configMap
-
emptyDir
- share space / state across containers in same pod
- containers can mount at different times
- pod crash --> data lost
- container crash --> ok
-
gitRepo
-
secret
- store on RAM
-
hostPath
Credit: https://www.reddit.com/user/__brennerm/
- Role
- Apply on namespace resources
- ClusterRole
- cluster-scoped resources (nodes,...)
- non-resources endpoint (/healthz)
- namespace resources across all namespaces
CustomResourceDefinitions themselves are non-namespaced and are available to all namespaces.
apiVersion: apiextensions.k8s.io/v1beta1
kind: CustomResourceDefinition
metadata:
# name must match the spec fields below, and be in the form: <plural>.<group>
name: crontabs.stable.example.com
spec:
# group name to use for REST API: /apis/<group>/<version>
group: stable.example.com
# version name to use for REST API: /apis/<group>/<version>
version: v1
# either Namespaced or Cluster
scope: Namespaced
names:
# plural name to be used in the URL: /apis/<group>/<version>/<plural>
plural: crontabs
# singular name to be used as an alias on the CLI and for display
singular: crontab
# kind is normally the CamelCased singular type. Your resource manifests use this.
kind: CronTab
# shortNames allow shorter string to match your resource on the CLI
shortNames:
- ct
# categories is a list of grouped resources the custom resource belongs to.
categories:
- all
validation:
# openAPIV3Schema is the schema for validating custom objects.
openAPIV3Schema:
properties:
spec:
properties:
cronSpec:
type: string
pattern: '^(\d+|\*)(/\d+)?(\s+(\d+|\*)(/\d+)?){4}$'
replicas:
type: integer
minimum: 1
maximum: 10
# subresources describes the subresources for custom resources.
subresources:
# status enables the status subresource.
status: {}
# scale enables the scale subresource.
scale:
# specReplicasPath defines the JSONPath inside of a custom resource that corresponds to Scale.Spec.Replicas.
specReplicasPath: .spec.replicas
# statusReplicasPath defines the JSONPath inside of a custom resource that corresponds to Scale.Status.Replicas.
statusReplicasPath: .status.replicas
# labelSelectorPath defines the JSONPath inside of a custom resource that corresponds to Scale.Status.Selector.
labelSelectorPath: .status.labelSelector# show current context
kubectl config current-context
# get specific resource
kubectl get (pod|svc|deployment|ingress) <resource-name>
# Get pod logs
kubectl logs -f <pod-name>
# Get nodes list
kubectl get no -o custom-columns=NAME:.metadata.name,AWS-INSTANCE:.spec.externalID,AGE:.metadata.creationTimestamp
# Run specific command | Drop to shell
kubectl exec -it <pod-name> <command>
# Describe specific resource
kubectl describe (pod|svc|deployment|ingress) <resource-name>
# Set context
kubectl config set-context $(kubectl config current-context) --namespace=<namespace-name>
# Run a test pod
kubectl run -it --rm --generator=run-pod/v1 --image=alpine:3.6 tuan-shell -- sh-
from @so0k link
-
access dashboard
# bash
kubectl -n kube-system port-forward $(kubectl get pods -n kube-system -o wide | grep dashboard | awk '{print $1}') 9090
# fish
kubectl -n kube-system port-forward (kubectl get pods -n kube-system -o wide | grep dashboard | awk '{print $1}') 9090From link
{
"kind": "List",
"items":[
{
"kind":"None",
"metadata":{"name":"127.0.0.1"},
"status":{
"capacity":{"cpu":"4"},
"addresses":[{"type": "LegacyHostIP", "address":"127.0.0.1"}]
}
},
{
"kind":"None",
"metadata":{"name":"127.0.0.2"},
"status":{
"capacity":{"cpu":"8"},
"addresses":[
{"type": "LegacyHostIP", "address":"127.0.0.2"},
{"type": "another", "address":"127.0.0.3"}
]
}
}
],
"users":[
{
"name": "myself",
"user": {}
},
{
"name": "e2e",
"user": {"username": "admin", "password": "secret"}
}
]
}| Function | Description | Example | Result |
|---|---|---|---|
| text | the plain text | kind is {.kind} | kind is List |
| @ | the current object | {@} | the same as input |
| . or [] | child operator | {.kind} or {['kind']} | List |
| .. | recursive descent | {..name} | 127.0.0.1 127.0.0.2 myself e2e |
| * | wildcard. Get all objects | {.items[*].metadata.name} | [127.0.0.1 127.0.0.2] |
| [start:end :step] | subscript operator | {.users[0].name} | myself |
| [,] | union operator | {.items[*]['metadata.name', 'status.capacity']} | 127.0.0.1 127.0.0.2 map[cpu:4] map[cpu:8] |
| ?() | filter | {.users[?(@.name=="e2e")].user.password} | secret |
| range, end | iterate list | {range .items[*]}[{.metadata.name}, {.status.capacity}] {end} | [127.0.0.1, map[cpu:4]] [127.0.0.2, map[cpu:8]] |
| '' | quote interpreted string | {range .items[*]}{.metadata.name}{'\t'}{end} | 127.0.0.1 127.0.0.2 |
Below are some examples using jsonpath:
$ kubectl get pods -o json
$ kubectl get pods -o=jsonpath='{@}'
$ kubectl get pods -o=jsonpath='{.items[0]}'
$ kubectl get pods -o=jsonpath='{.items[0].metadata.name}'
$ kubectl get pods -o=jsonpath='{range .items[*]}{.metadata.name}{"\t"}{.status.startTime}{"\n"}{end}'The CPU resource is measured in cpu units. One cpu, in Kubernetes, is equivalent to:
- 1 AWS vCPU
- 1 GCP Core
- 1 Azure vCore
- 1 Hyperthread on a bare-metal Intel processor with Hyperthreading
The memory resource is measured in bytes. You can express memory as a plain integer or a fixed-point integer with one of these suffixes: E, P, T, G, M, K, Ei, Pi, Ti, Gi, Mi, Ki. For example, the following represent approximately the same value:
128974848, 129e6, 129M , 123Mi
- External service without selector (access with
external-database.svc.default.clusterendpoint)
kind: Service
apiVersion: v1
metadata:
name: external-database
spec:
type: ExternalName
externalName: "database.company.com- external service with IP only
kind: Service
apiVersion: v1
metadata:
name: external-ip-database
---
kind: Endpoints
apiVersion: v1
metadata:
name: external-ip-database
subsets:
- addresses:
- ip: 192.168.0.1
ports:
- port: 3306The following information is available to containers through environment variables and downwardAPI volumes:
Information available via fieldRef:
- spec.nodeName - the node’s name
- status.hostIP - the node’s IP
- metadata.name - the pod’s name
- metadata.namespace - the pod’s namespace
- status.podIP - the pod’s IP address
- spec.serviceAccountName - the pod’s service account name
- metadata.uid - the pod’s UID
- metadata.labels[''] - the value of the pod’s label (for example, metadata.labels['mylabel']); available in Kubernetes 1.9+
- metadata.annotations[''] - the value of the pod’s annotation (for example, metadata.annotations['myannotation']); available in Kubernetes 1.9+
- Information available via resourceFieldRef:
- A Container’s CPU limit
- A Container’s CPU request
- A Container’s memory limit
- A Container’s memory request
In addition, the following information is available through downwardAPI volume fieldRef:
- metadata.labels - all of the pod’s labels, formatted as label-key="escaped-label-value" with one label per line
- metadata.annotations - all of the pod’s annotations, formatted as annotation-key="escaped-annotation-value" with one annotation per line
See link
- Marking pod as critical when using Rescheduler. To be considered critical, the pod has to:
- Run in the
kube-systemnamespace (configurable via flag) - Have the
scheduler.alpha.kubernetes.io/critical-podannotation set to empty string - Have the PodSpec’s tolerations field set to
[{"key":"CriticalAddonsOnly", "operator":"Exists"}].
- Run in the
The first one marks a pod a critical. The second one is required by Rescheduler algorithm.
- Marking pod as critical when priorites are enabled. To be considered critical, the pod has to:
- Run in the
kube-systemnamespace (configurable via flag) - Have the priorityClass set as
system-cluster-criticalorsystem-node-critical, the latter being the highest for entire cluster scheduler.alpha.kubernetes.io/critical-podannotation set to empty string(This will be deprecated too).
- Run in the
env:
- name: MESSAGE
value: "hello world"
command: ["/bin/echo"]
args: ["$(MESSAGE)"]
sample deployment: