I am fairly sure this isn't possible, but I wanted to check.
I am using Kubernetes stateful sets, so my hosts get obvious hostnames.
I'd like them to provision a hostPath mount that is mapped to their hostname.
An example helm chart that I'm using might look like this:
apiVersion: apps/v1beta1
kind: StatefulSet
metadata:
name: app
namespace: '{{ .Values.name }}'
labels:
chart: '{{ .Chart.Name }}-{{ .Chart.Version | replace "+" "_" }}'
spec:
serviceName: "app"
replicas: {{ .Values.replicaCount }}
template:
metadata:
labels:
app: app
spec:
terminationGracePeriodSeconds: 30
containers:
- name: {{ .Chart.Name }}
image: "{{ .Values.image.repository }}/{{ .Values.image.version}}"
imagePullPolicy: '{{ .Values.image.pullPolicy }}'
env:
- name: POD_NAME
valueFrom:
fieldRef:
fieldPath: metadata.name
ports:
- containerPort: {{ .Values.baseport | add 80 }}
name: app
volumeMounts:
- mountPath: /NAS/$(POD_NAME)
name: store
readOnly: true
volumes:
- name: store
hostPath:
path: /NAS/$(POD_NAME)
Essentially, instead of hardcoding a volume, I'd like to have some kind of dynamic variable as the path. I don't mind using helm or the downward API for this, but ideally it would work when I scale the stateful set outwards.
Is there any way of doing this? All my docs reading seems to think it's not... :(
Related
I have a ConfigMap with multiple files, and want to add these files to an already existing directory. But the tricky part here is, the filenames(keys) can change. So I can't try to mount them individually using subPath.
Is there any way this can be achieved from Deployment manifest?
Configmap:
config-files-configmap
└── newFile1.yml
└── newFile2.yml
Existing directory after adding files from configmap:
config/
└── existingFile1.yml
└── existingFile2.yml
└── newFile1.yml
└── newFile2.yml
PS: I have tried mounting the configmap as directory, which will override existing contents of the directory.
Thanks
You can use the init container with configmap as a volume mount.
Not sure about the actual deployment architecture.
i would suggest injecting the configmap files to another directory and copying and pasting at starting of the main container.
Using life cycle hook of POD of init container.
As we can not go with subpath, this one option i am seeing as of now.
Example helm template from RabbitMQ
apiVersion: apps/v1
kind: StatefulSet
metadata:
name: {{ .Release.Name }}-rabbitmq
labels: &RabbitMQDeploymentLabels
app.kubernetes.io/name: {{ .Release.Name }}
app.kubernetes.io/component: rabbitmq-server
spec:
selector:
matchLabels: *RabbitMQDeploymentLabels
serviceName: {{ .Release.Name }}-rabbitmq-discovery
replicas: {{ .Values.rabbitmq.replicas }}
updateStrategy:
# https://www.rabbitmq.com/upgrade.html
# https://cloud.google.com/kubernetes-engine/docs/how-to/updating-apps
type: RollingUpdate
template:
metadata:
labels: *RabbitMQDeploymentLabels
spec:
serviceAccountName: {{ .Values.rabbitmq.serviceAccount }}
terminationGracePeriodSeconds: 180
initContainers:
# This init container copies the config files from read-only ConfigMap to writable location.
- name: copy-rabbitmq-config
image: {{ .Values.rabbitmq.initImage }}
imagePullPolicy: Always
command:
- /bin/bash
- -euc
- |
# Remove cached erlang cookie since we are always providing it,
# that opens the way to recreate the application and access to existing data
# as a new erlang will be regenerated again.
echo ${RABBITMQ_ERLANG_COOKIE} > /var/lib/rabbitmq/.erlang.cookie
chmod 600 /var/lib/rabbitmq/.erlang.cookie
# Copy the mounted configuration to both places.
cp /rabbitmqconfig/rabbitmq.conf /etc/rabbitmq/rabbitmq.conf
# Change permission to allow to add more configurations via variables
chown :999 /etc/rabbitmq/rabbitmq.conf
chmod 660 /etc/rabbitmq/rabbitmq.conf
cp /rabbitmqconfig/enabled_plugins /etc/rabbitmq/enabled_plugins
volumeMounts:
- name: configmap
mountPath: /rabbitmqconfig
- name: config
mountPath: /etc/rabbitmq
- name: {{ .Release.Name }}-rabbitmq-pvc
mountPath: /var/lib/rabbitmq
env:
- name: RABBITMQ_ERLANG_COOKIE
valueFrom:
secretKeyRef:
name: {{ .Release.Name }}-rabbitmq-secret
key: rabbitmq-erlang-cookie
containers:
- name: rabbitmq
image: "{{ .Values.rabbitmq.image.repo }}:{{ .Values.rabbitmq.image.tag }}"
imagePullPolicy: Always
resources:
requests:
cpu: 100m
memory: 100Mi
env:
- name: MY_POD_NAME
valueFrom:
fieldRef:
fieldPath: metadata.name
- name: RABBITMQ_USE_LONGNAME
value: 'true'
- name: RABBITMQ_NODENAME
value: 'rabbit#$(MY_POD_NAME).{{ .Release.Name }}-rabbitmq-discovery.{{ .Release.Namespace }}.svc.cluster.local'
- name: K8S_SERVICE_NAME
value: '{{ .Release.Name }}-rabbitmq-discovery'
- name: K8S_HOSTNAME_SUFFIX
value: '.{{ .Release.Name }}-rabbitmq-discovery.{{ .Release.Namespace }}.svc.cluster.local'
# User name to create when RabbitMQ creates a new database from scratch.
- name: RABBITMQ_DEFAULT_USER
value: '{{ .Values.rabbitmq.user }}'
# Password for the default user.
- name: RABBITMQ_DEFAULT_PASS
valueFrom:
secretKeyRef:
name: {{ .Release.Name }}-rabbitmq-secret
key: rabbitmq-pass
ports:
- name: clustering
containerPort: 25672
- name: amqp
containerPort: 5672
- name: amqp-ssl
containerPort: 5671
- name: prometheus
containerPort: 15692
- name: http
containerPort: 15672
volumeMounts:
- name: config
mountPath: /etc/rabbitmq
- name: {{ .Release.Name }}-rabbitmq-pvc
mountPath: /var/lib/rabbitmq
livenessProbe:
exec:
command:
- rabbitmqctl
- status
initialDelaySeconds: 60
timeoutSeconds: 30
readinessProbe:
exec:
command:
- rabbitmqctl
- status
initialDelaySeconds: 20
timeoutSeconds: 30
lifecycle:
postStart:
exec:
command:
- /bin/bash
- -c
- |
# Wait for the RabbitMQ to be ready.
until rabbitmqctl node_health_check; do
sleep 5
done
# By default, RabbitMQ does not have Highly Available policies enabled,
# using the following command to enable it.
rabbitmqctl set_policy ha-all "." '{"ha-mode":"all", "ha-sync-mode":"automatic"}' --apply-to all --priority 0
{{ if .Values.metrics.exporter.enabled }}
- name: prometheus-to-sd
image: {{ .Values.metrics.image }}
ports:
- name: profiler
containerPort: 6060
command:
- /monitor
- --stackdriver-prefix=custom.googleapis.com
- --source=rabbitmq:http://localhost:15692/metrics
- --pod-id=$(POD_NAME)
- --namespace-id=$(POD_NAMESPACE)
- --monitored-resource-type-prefix=k8s_
env:
- name: POD_NAME
valueFrom:
fieldRef:
fieldPath: metadata.name
- name: POD_NAMESPACE
valueFrom:
fieldRef:
fieldPath: metadata.namespace
{{ end }}
volumes:
- name: configmap
configMap:
name: {{ .Release.Name }}-rabbitmq-config
items:
- key: rabbitmq.conf
path: rabbitmq.conf
- key: enabled_plugins
path: enabled_plugins
- name: config
emptyDir: {}
volumeClaimTemplates:
- metadata:
name: {{ .Release.Name }}-rabbitmq-pvc
labels: *RabbitMQDeploymentLabels
spec:
accessModes:
- ReadWriteOnce
storageClassName: {{ .Values.rabbitmq.persistence.storageClass }}
resources:
requests:
storage: {{ .Values.rabbitmq.persistence.size }}
Example reference : https://github.com/GoogleCloudPlatform/click-to-deploy/blob/master/k8s/rabbitmq/chart/rabbitmq/templates/statefulset.yaml
I have a deployment .yaml file that basically create a pod with mariadb, as follows
apiVersion: apps/v1
kind: Deployment
metadata:
name: {{ .Release.Name }}-pod
spec:
replicas: 1
selector:
matchLabels:
pod: {{ .Release.Name }}-pod
strategy:
type: Recreate
template:
metadata:
labels:
pod: {{ .Release.Name }}-pod
spec:
containers:
- env:
- name: MYSQL_ROOT_PASSWORD
value: {{ .Values.db.password }}
image: {{ .Values.image.repository }}
name: {{ .Release.Name }}
ports:
- containerPort: 3306
resources:
requests:
memory: 2048Mi
cpu: 0.5
limits:
memory: 4096Mi
cpu: 1
volumeMounts:
- mountPath: /var/lib/mysql
name: dbsvr-claim
- mountPath: /etc/mysql/conf.d/my.cnf
name: conf
subPath: my.cnf
- mountPath: /docker-entrypoint-initdb.d/init.sql
name: conf
subPath: init.sql
restartPolicy: Always
volumes:
- name: dbsvr-claim
persistentVolumeClaim:
claimName: {{ .Release.Name }}-claim
- name: conf
configMap:
name: {{ .Release.Name }}-configmap
status: {}
Upon success on
helm install abc ./abc/ -f values.yaml
I have a job that generates a mysqldump backup file and it completes successfully (just showing the relevant code)
apiVersion: batch/v1
kind: Job
metadata:
name: {{ .Release.Name }}-job
spec:
template:
metadata:
name: {{ .Release.Name }}-job
spec:
containers:
- name: {{ .Release.Name }}-dbload
image: {{ .Values.image.repositoryRoot }}/{{.Values.image.imageName}}
command: ["/bin/sh", "-c"]
args:
- mysqldump -p$(PWD) -h{{.Values.db.source}} -u$(USER) --databases xyz > $(FILE);
echo "done!";
imagePullPolicy: Always
# Do not restart containers after they exit
restartPolicy: Never
So, here's my question. Is there a way to automatically start the job after the helm install abc ./ -f values.yaml finishes with success?
you can use kubectl wait -h command to execute job when the condition=Ready for the deployment.
Here the article wait-for-condition demonstrate quite similar situation
I'm looking for a possible way to reference the secrets in my deployment.yaml (1 liner)
Currently I'm using the
containers:
- name: {{ template "myapp.name" . }}
image: "{{ .Values.image.repository }}:{{ .Values.image.tag }}"
imagePullPolicy: Always
env:
- name: COUCHDB_USER
valueFrom:
secretKeyRef:
name: {{ .Release.Name }}-secrets
key: COUCHDB_USER
- name: COUCHDB_PASSWORD
valueFrom:
secretKeyRef:
name: {{ .Release.Name }}-secrets
key: COUCHDB_PASSWORD
With the minimal modification possible, I want to achieve something like this:
containers:
- name: {{ template "myapp.name" . }}
image: "{{ .Values.image.repository }}:{{ .Values.image.tag }}"
imagePullPolicy: Always
env:
- name: COUCHDB_URL
value: http://${COUCHDB_USER}:${COUCHDB_PASSWORD}#{{ .Release.Name }}-couchdb:5984
Just carious if I can do this in 1 step in during the deployment, instead of passing 2 env vars and parse them in my application.
I am not seeing any way to achieve it without setting COUCHDB_USER and COUCHDB_PASSWORD in container env.
One workaround is, you can specify your secret in container.EnvFrom and all your secret keys will be converted to Environment variables. then, You can use those environment variables to create your composite env (ie, COUCHDB_URL).
FYI, To create env from another env in kubernetes, () is used. Curly braces {} won't work at this very moment.
One sample is,
apiVersion: v1
kind: Secret
metadata:
name: mysecret
type: Opaque
data:
COUCHDB_USER: YWRtaW4=
COUCHDB_PASSWORD: MWYyZDFlMmU2N2Rm
---
apiVersion: v1
kind: Pod
metadata:
name: secret-env-pod
spec:
containers:
- name: mycontainer
image: redis
envFrom:
- secretRef:
name: mysecret
env:
- name: COUCHDB_URL
value: http://$(COUCHDB_USER):$(COUCHDB_PASSWORD)rest-of-the-url
You can confirm, the output by,
$ kubectl exec -it secret-env-pod bash
root#secret-env-pod:/data# env | grep COUCHDB
COUCHDB_URL=http://admin:1f2d1e2e67dfrest-of-the-url
COUCHDB_PASSWORD=1f2d1e2e67df
COUCHDB_USER=admin
In your case, the yaml for container is:
containers:
- name: {{ template "myapp.name" . }}
image: "{{ .Values.image.repository }}:{{ .Values.image.tag }}"
imagePullPolicy: Always
envFrom:
- secretRef:
name: {{ .Release.Name }}-secrets
env:
- name: COUCHDB_URL
value: http://$(COUCHDB_USER):$(COUCHDB_PASSWORD)#{{ .Release.Name }}-couchdb:5984
I am using helm with kubernetes on google cloud platform.
i get the following error for my postgres deployment:
SchedulerPredicates failed due to PersistentVolumeClaim is not bound
it looks like it cant connect to the persistent storage but i don't understand why because the persistent storage loaded fine.
i have tried deleting the helm release completely, then on google-cloud-console > compute-engine > disks; i have deleted all persistent disk. and finally tried to install from the helm chart, but the postgres deployment still doesnt connect to the PVC.
my database configuration:
{{- $serviceName := "db-service" -}}
{{- $deploymentName := "db-deployment" -}}
{{- $pvcName := "db-disk-claim" -}}
{{- $pvName := "db-disk" -}}
apiVersion: v1
kind: Service
metadata:
name: {{ $serviceName }}
labels:
name: {{ $serviceName }}
env: production
spec:
type: LoadBalancer
ports:
- port: 5432
targetPort: 5432
protocol: TCP
name: http
selector:
name: {{ $deploymentName }}
---
apiVersion: apps/v1beta1
kind: Deployment
metadata:
name: {{ $deploymentName }}
labels:
name: {{ $deploymentName }}
env: production
spec:
replicas: 1
template:
metadata:
labels:
name: {{ $deploymentName }}
env: production
spec:
containers:
- name: postgres-database
image: postgres:alpine
imagePullPolicy: Always
env:
- name: POSTGRES_USER
value: test-user
- name: POSTGRES_PASSWORD
value: test-password
- name: POSTGRES_DB
value: test_db
- name: PGDATA
value: /var/lib/postgresql/data/pgdata
ports:
- containerPort: 5432
volumeMounts:
- mountPath: "/var/lib/postgresql/data/pgdata"
name: {{ $pvcName }}
volumes:
- name: {{ $pvcName }}
persistentVolumeClaim:
claimName: {{ $pvcName }}
---
apiVersion: v1
kind: PersistentVolumeClaim
metadata:
name: {{ $pvcName }}
spec:
accessModes:
- ReadWriteOnce
resources:
requests:
storage: 1Gi
selector:
matchLabels:
name: {{ $pvName }}
env: production
---
apiVersion: v1
kind: PersistentVolume
metadata:
name: {{ .Values.gcePersistentDisk }}
labels:
name: {{ $pvName }}
env: production
annotations:
volume.beta.kubernetes.io/mount-options: "discard"
spec:
capacity:
storage: 10Gi
accessModes:
- ReadWriteOnce
gcePersistentDisk:
fsType: "ext4"
pdName: {{ .Values.gcePersistentDisk }}
is this config for kubenetes correct? i have read the documentation and it looks like this should work. i'm new to Kubernetes and helm so any advice is appreciated.
EDIT:
i have added a PersistentVolume and linked it to the PersistentVolumeClaim to see if that helps, but it seems that when i do this, the PersistentVolumeClaim status becomes stuck in "pending" (resulting in the same issue as before).
You don't have a bound PV for this claim. What storage you use for this claim. You need to mention it in the PVC file
I'm running a kubernetes 1.6.2 cluster across three nodes in different zones in GKE and I'm trying to deploy my statefulset where each pod in the statefulset gets a PV attached to it. The problem is that kubernetes is creating the PVs in the one zone where I don't have a node!
$ kubectl describe node gke-multi-consul-default-pool-747c9378-zls3|grep 'zone=us-central1'
failure-domain.beta.kubernetes.io/zone=us-central1-a
$ kubectl describe node gke-multi-consul-default-pool-7e987593-qjtt|grep 'zone=us-central1'
failure-domain.beta.kubernetes.io/zone=us-central1-f
$ kubectl describe node gke-multi-consul-default-pool-8e9199ea-91pj|grep 'zone=us-central1'
failure-domain.beta.kubernetes.io/zone=us-central1-c
$ kubectl describe pv pvc-3f668058-2c2a-11e7-a7cd-42010a8001e2|grep 'zone=us-central1'
failure-domain.beta.kubernetes.io/zone=us-central1-b
I'm using the standard storageclass which has no default zone set:
$ kubectl describe storageclass standard
Name: standard
IsDefaultClass: Yes
Annotations: storageclass.beta.kubernetes.io/is-default-class=true
Provisioner: kubernetes.io/gce-pd
Parameters: type=pd-standard
Events: <none>
So I thought that the scheduler would automatically provision the volumes in a zone where a cluster node existed, but it doesn't seem to be doing that.
For reference, here is the yaml for my statefulset:
apiVersion: apps/v1beta1
kind: StatefulSet
metadata:
name: "{{ template "fullname" . }}"
labels:
heritage: {{.Release.Service | quote }}
release: {{.Release.Name | quote }}
chart: "{{.Chart.Name}}-{{.Chart.Version}}"
component: "{{.Release.Name}}-{{.Values.Component}}"
spec:
serviceName: "{{ template "fullname" . }}"
replicas: {{default 3 .Values.Replicas}}
template:
metadata:
name: "{{ template "fullname" . }}"
labels:
heritage: {{.Release.Service | quote }}
release: {{.Release.Name | quote }}
chart: "{{.Chart.Name}}-{{.Chart.Version}}"
component: "{{.Release.Name}}-{{.Values.Component}}"
app: "consul"
annotations:
pod.alpha.kubernetes.io/initialized: "true"
spec:
securityContext:
fsGroup: 1000
containers:
- name: "{{ template "fullname" . }}"
image: "{{.Values.Image}}:{{.Values.ImageTag}}"
imagePullPolicy: "{{.Values.ImagePullPolicy}}"
ports:
- name: http
containerPort: {{.Values.HttpPort}}
- name: rpc
containerPort: {{.Values.RpcPort}}
- name: serflan-tcp
protocol: "TCP"
containerPort: {{.Values.SerflanPort}}
- name: serflan-udp
protocol: "UDP"
containerPort: {{.Values.SerflanUdpPort}}
- name: serfwan-tcp
protocol: "TCP"
containerPort: {{.Values.SerfwanPort}}
- name: serfwan-udp
protocol: "UDP"
containerPort: {{.Values.SerfwanUdpPort}}
- name: server
containerPort: {{.Values.ServerPort}}
- name: consuldns
containerPort: {{.Values.ConsulDnsPort}}
resources:
requests:
cpu: "{{.Values.Cpu}}"
memory: "{{.Values.Memory}}"
env:
- name: INITIAL_CLUSTER_SIZE
value: {{ default 3 .Values.Replicas | quote }}
- name: STATEFULSET_NAME
value: "{{ template "fullname" . }}"
- name: POD_IP
valueFrom:
fieldRef:
fieldPath: status.podIP
- name: STATEFULSET_NAMESPACE
valueFrom:
fieldRef:
fieldPath: metadata.namespace
volumeMounts:
- name: datadir
mountPath: /var/lib/consul
- name: gossip-key
mountPath: /etc/secrets
readOnly: true
- name: config
mountPath: /etc/consul
- name: tls
mountPath: /etc/tls
lifecycle:
preStop:
exec:
command:
- /bin/sh
- -c
- consul leave
livenessProbe:
exec:
command:
- consul
- members
initialDelaySeconds: 300
timeoutSeconds: 5
command:
- "/bin/sh"
- "-ec"
- "/tmp/consul-start.sh"
volumes:
- name: config
configMap:
name: consul
- name: gossip-key
secret:
secretName: {{ template "fullname" . }}-gossip-key
- name: tls
secret:
secretName: consul
volumeClaimTemplates:
- metadata:
name: datadir
annotations:
{{- if .Values.StorageClass }}
volume.beta.kubernetes.io/storage-class: {{.Values.StorageClass | quote}}
{{- else }}
volume.alpha.kubernetes.io/storage-class: default
{{- end }}
spec:
accessModes:
- "ReadWriteOnce"
resources:
requests:
# upstream recommended max is 700M
storage: "{{.Values.Storage}}"
There is a bug open for this issue here.
The workaround in the meantime is to set the zones parameter in your StorageClass to specify the exact zones where your Kubernetes cluster has nodes. Here is an example.
Answer from the Kubernetes documentation about Persistent Volumes: https://kubernetes.io/docs/concepts/storage/persistent-volumes/#gce
zone: GCE zone. If not specified, a random zone in the same region as controller-manager will be chosen.
I guess your controller manager is in region us-central-1, so any zone can be choosen from that region, in your case I guess the only zone that is not covered is us-central-1b, so you have to start a node there as well, or set the zone in the StorageClass resource.
You could create storage classes for each zone, then a PV/PVC may specify that storage class. Your stateful sets/deployments could be set up to target a specific node via nodeSelector so they always get scheduled on a node in a specific zone (see built-in node labels)
storage_class.yml
kind: StorageClass
apiVersion: storage.k8s.io/v1
metadata:
name: us-central-1a
provisioner: kubernetes.io/gce-pd
parameters:
type: pd-standard
zone: us-central1-a
persistent_volume.yml
apiVersion: v1
kind: PersistentVolume
metadata:
name: some-volume
spec:
capacity:
storage: 5Gi
accessModes:
- ReadWriteOnce
storageClassName: us-central-1a
Note that you can use storageClassName in kubernetes 1.6, or otherwise the annotation volume.beta.kubernetes.io/storage-class should work too (however will deprecate in the future).