Trying to deploy a stateful jupyter notebook in Kubernetes, but not able to save the code written in a notebook, whenever the notebook pod is going down all the code is being deleted. I tried to use persistent volume but unable to achieve the expected result.
UPDATE
Changed mount path to "/home/jovyan" as jyputer saves the ipynb in this location. But now getting PermissionError: [Errno 13] Permission denied: '/home/jovyan/.local' while deploying the pod.
kind: Ingress
metadata:
name: jupyter-ingress
spec:
backend:
serviceName: jupyter-notebook-service
servicePort: 8888
---
kind: Service
apiVersion: v1
metadata:
name: jupyter-notebook-service
spec:
clusterIP: None
selector:
app: jupyter-notebook
ports:
- protocol: TCP
port: 8888
targetPort: 8888
---
apiVersion: apps/v1
kind: StatefulSet
metadata:
name: jupyter-notebook
labels:
app: jupyter-notebook
spec:
replicas: 1
serviceName: "jupyter-notebook-service"
selector:
matchLabels:
app: jupyter-notebook
template:
metadata:
labels:
app: jupyter-notebook
spec:
serviceAccountName: dsx-spark
volumes:
- name: jupyter-pv-storage
persistentVolumeClaim:
claimName: jupyter-pv-claim
containers:
- name: minimal-notebook
image: jupyter/pyspark-notebook:latest
ports:
- containerPort: 8888
command: ["start-notebook.sh"]
args: ["--NotebookApp.token=''"]
volumeMounts:
- mountPath: "/home/jovyan"
name: jupyter-pv-storage
---
kind: PersistentVolumeClaim
apiVersion: v1
metadata:
name: jupyter-pv-claim
spec:
storageClassName: jupyter-pv-storage
accessModes:
- ReadWriteOnce
resources:
requests:
storage: 1Gi
---
---
apiVersion: v1
kind: PersistentVolume
metadata:
name: jupyter-pv-volume
labels:
type: local
spec:
storageClassName: jupyter-pv-storage
capacity:
storage: 1Gi
accessModes:
- ReadWriteOnce
hostPath:
path: "/home/jovyan"
---
apiVersion: storage.k8s.io/v1
kind: StorageClass
metadata:
name: jupyternotebook-pv-storage
annotations:
storageclass.kubernetes.io/is-default-class: "true"
labels:
provisioner: kubernetes.io/vsphere-volume
parameters:
diskformat: zeroedthick
```
The Pod with jupyter notebook is non-root user so unable to mount the so we are
using initContainers to change user/permission of the Persistent Volume Claim before creating the Pod.
kind: StatefulSet
metadata:
name: jupyter-notebook
labels:
app: jupyter-notebook
spec:
replicas: 1
serviceName: "jupyter-notebook-service"
selector:
matchLabels:
app: jupyter-notebook
template:
metadata:
labels:
app: jupyter-notebook
spec:
serviceAccountName: dsx-spark
volumes:
- name: ci-jupyter-storage-def
persistentVolumeClaim:
claimName: my-jupyter-pv-claim
containers:
- name: minimal-notebook
image: jupyter/pyspark-notebook:latest
ports:
- containerPort: 8888
command: ["start-notebook.sh"]
args: ["--NotebookApp.token=''"]
volumeMounts:
- mountPath: "/home/jovyan/work"
name: ci-jupyter-storage-def
initContainers:
- name: jupyter-data-permission-fix
image: busybox
command: ["/bin/chmod","-R","777", "/data"]
volumeMounts:
- name: ci-jupyter-storage-def
mountPath: /data```
As I have already mentioned in the comments you need to make sure that:
The storage for a given Pod must either be provisioned by a PersistentVolume Provisioner based on the requested storage class, or pre-provisioned by an admin. The volumeClaimTemplates will provide stable storage using PersistentVolumes. PersistentVolumes associated with the Pods’ PersistentVolume Claims are not deleted when the Pods, or StatefulSet are deleted.
The container is running as a user that has the permissions to access that volume. It can be done by changing the permissions to 777 or as you already noticed by using a proper initContainers.
Related
I have a Raspberry Pi Cluster consisting of 1-Master 20-Nodes:
192.168.0.92 (Master)
192.168.0.112 (Node w/ USB Drive)
I mounted a USB drive to /media/hdd & set a label - purpose=volume to it.
Using the following I was able to setup a NFS server:
apiVersion: v1
kind: Namespace
metadata:
name: storage
labels:
app: storage
---
apiVersion: v1
kind: PersistentVolume
metadata:
name: local-pv
namespace: storage
spec:
capacity:
storage: 3.5Ti
accessModes:
- ReadWriteOnce
persistentVolumeReclaimPolicy: Retain
storageClassName: local-storage
local:
path: /media/hdd
nodeAffinity:
required:
nodeSelectorTerms:
- matchExpressions:
- key: purpose
operator: In
values:
- volume
---
kind: PersistentVolumeClaim
apiVersion: v1
metadata:
name: local-claim
namespace: storage
spec:
accessModes:
- ReadWriteOnce
storageClassName: local-storage
resources:
requests:
storage: 3Ti
---
apiVersion: apps/v1
kind: Deployment
metadata:
name: nfs-server
namespace: storage
labels:
app: nfs-server
spec:
replicas: 1
selector:
matchLabels:
app: nfs-server
template:
metadata:
labels:
app: nfs-server
name: nfs-server
spec:
containers:
- name: nfs-server
image: itsthenetwork/nfs-server-alpine:11-arm
env:
- name: SHARED_DIRECTORY
value: /exports
ports:
- name: nfs
containerPort: 2049
- name: mountd
containerPort: 20048
- name: rpcbind
containerPort: 111
securityContext:
privileged: true
volumeMounts:
- mountPath: /exports
name: mypvc
volumes:
- name: mypvc
persistentVolumeClaim:
claimName: local-claim
nodeSelector:
purpose: volume
---
kind: Service
apiVersion: v1
metadata:
name: nfs-server
namespace: storage
spec:
ports:
- name: nfs
port: 2049
- name: mountd
port: 20048
- name: rpcbind
port: 111
clusterIP: 10.96.0.11
selector:
app: nfs-server
And I was even able to make a persistent volume with this:
apiVersion: v1
kind: PersistentVolume
metadata:
name: mysql-nfs-volume
labels:
directory: mysql
spec:
capacity:
storage: 200Gi
volumeMode: Filesystem
accessModes:
- ReadWriteOnce
persistentVolumeReclaimPolicy: Retain
storageClassName: slow
nfs:
path: /mysql
server: 10.244.19.5
---
apiVersion: v1
kind: PersistentVolumeClaim
metadata:
name: mysql-nfs-claim
spec:
storageClassName: slow
accessModes:
- ReadWriteOnce
resources:
requests:
storage: 100Gi
selector:
matchLabels:
directory: mysql
But when I try to use the volume like so:
apiVersion: v1
kind: Service
metadata:
name: wordpress-mysql
labels:
app: wordpress
spec:
ports:
- port: 3306
selector:
app: wordpress
tier: mysql
clusterIP: None
---
apiVersion: apps/v1
kind: Deployment
metadata:
name: wordpress-mysql
labels:
app: wordpress
spec:
selector:
matchLabels:
app: wordpress
tier: mysql
strategy:
type: Recreate
template:
metadata:
labels:
app: wordpress
tier: mysql
spec:
containers:
- image: mysql:5.6
name: mysql
env:
- name: MYSQL_ROOT_PASSWORD
valueFrom:
secretKeyRef:
name: mysql-pass
key: password
ports:
- containerPort: 3306
name: mysql
volumeMounts:
- name: mysql-persistent-storage
mountPath: /var/lib/mysql
volumes:
- name: mysql-persistent-storage
persistentVolumeClaim:
claimName: mysql-nfs-claim
I get NFS version transport protocol not supported error.
When seeing mount.nfs: requested NFS version or transport protocol is not supported error there are three main reasons:
NFS services are not running on NFS server
NFS utils not installed on the client
NFS service hung on NFS server
According tho this artice there are three solutions to resolve the problem with your error.
First one:
Login to the NFS server and check the NFS services status. If the following command
service nfs status returns an information that NFS services are stopped on the server - just start them using service nfs start. To mount NFS share on the client use the same command.
Second one:
If after trying first solution your problem isn't resolved
try installing package nfs-utils on your server.
Third one:
Open file /etc/sysconfig/nfs and try to check below parameters
# Turn off v4 protocol support
#RPCNFSDARGS="-N 4"
# Turn off v2 and v3 protocol support
#RPCNFSDARGS="-N 2 -N 3"
Removing hash from RPCNFSDARGS lines will turn off specific version support. This way clients with mentioned NFS versions won’t be able to connect to the NFS server for mounting share. If you have any of it enabled, try disabling it and mounting at the client after the NFS server service restarts.
Could someone help me please and point me what configuration should I be doing for my use-case?
I'm building a development k8s cluster and one of the steps is to generate security files (private keys) that are generated in a number of pods during deployment (let's say for a simple setup I have 6 pods that each build their own security keys). I need to have access to all these files, also they must be persistent after the pod goes down.
I'm trying to figure out now how to set up it locally for internal testing. From what I understand Local PersistentVolumes only allow 1:1 with PersistentVolumeClaims, so I would have to create a separate PersistentVolume and PersistentVolumeClaim for each pod that get's configured. I would prefer to void this and use one PersistentVolume for all.
Could someone be so nice and help me or point me to the right setup that should be used?
-- Update: 26/11/2020
So this is my setup:
apiVersion: apps/v1
kind: Deployment
metadata:
name: hlf-nfs--server
spec:
replicas: 1
selector:
matchLabels:
app: hlf-nfs--server
template:
metadata:
labels:
app: hlf-nfs--server
spec:
containers:
- name: hlf-nfs--server
image: itsthenetwork/nfs-server-alpine:12
ports:
- containerPort: 2049
name: tcp
- containerPort: 111
name: udp
securityContext:
privileged: true
env:
- name: SHARED_DIRECTORY
value: "/opt/k8s-pods/data"
volumeMounts:
- name: pvc
mountPath: /opt/k8s-pods/data
volumes:
- name: pvc
persistentVolumeClaim:
claimName: shared-nfs-pvc
apiVersion: v1
kind: Service
metadata:
name: hlf-nfs--server
labels:
name: hlf-nfs--server
spec:
type: ClusterIP
selector:
app: hlf-nfs--server
ports:
- name: tcp-2049
port: 2049
protocol: TCP
- name: udp-111
port: 111
protocol: UDP
apiVersion: v1
kind: PersistentVolumeClaim
metadata:
name: shared-nfs-pvc
spec:
accessModes:
- ReadWriteMany
storageClassName: nfs
resources:
requests:
storage: 1Gi
These three are being created at once, after that, I'm reading the IP of the service and adding it to the last one:
apiVersion: v1
kind: PersistentVolume
metadata:
name: shared-nfs-pv
spec:
capacity:
storage: 100Gi
accessModes:
- ReadWriteMany
nfs:
path: /opt/k8s-pods/data
server: <<-- IP from `kubectl get svc -l name=hlf-nfs--server`
The problem I'm getting and trying to resolve is that the PVC does not get bound with the PV and the deployment keeps in READY mode.
Did I miss anything?
You can create a NFS and have the pods use NFS volume. Here is the manifest file to create such in-cluster NFS server (make sure you modify STORAGE_CLASS and the other variables below):
export NFS_NAME="nfs-share"
export NFS_SIZE="10Gi"
export NFS_IMAGE="itsthenetwork/nfs-server-alpine:12"
export STORAGE_CLASS="thin-disk"
kubectl apply -f - <<EOF
apiVersion: v1
kind: Service
metadata:
name: ${NFS_NAME}
labels:
app.kubernetes.io/name: nfs-server
app.kubernetes.io/instance: ${NFS_NAME}
spec:
ports:
- name: tcp-2049
port: 2049
protocol: TCP
- name: udp-111
port: 111
protocol: UDP
selector:
app.kubernetes.io/name: nfs-server
app.kubernetes.io/instance: ${NFS_NAME}
---
apiVersion: v1
kind: PersistentVolumeClaim
metadata:
labels:
app.kubernetes.io/name: nfs-server
app.kubernetes.io/instance: ${NFS_NAME}
name: ${NFS_NAME}
spec:
accessModes:
- ReadWriteOnce
resources:
requests:
storage: 10Gi
storageClassName: $STORAGE_CLASS
volumeMode: Filesystem
---
apiVersion: apps/v1
kind: Deployment
metadata:
name: ${NFS_NAME}
labels:
app.kubernetes.io/name: nfs-server
app.kubernetes.io/instance: ${NFS_NAME}
spec:
replicas: 1
selector:
matchLabels:
app.kubernetes.io/name: nfs-server
app.kubernetes.io/instance: ${NFS_NAME}
template:
metadata:
labels:
app.kubernetes.io/name: nfs-server
app.kubernetes.io/instance: ${NFS_NAME}
spec:
containers:
- name: nfs-server
image: ${NFS_IMAGE}
ports:
- containerPort: 2049
name: tcp
- containerPort: 111
name: udp
securityContext:
privileged: true
env:
- name: SHARED_DIRECTORY
value: /nfsshare
volumeMounts:
- name: pvc
mountPath: /nfsshare
volumes:
- name: pvc
persistentVolumeClaim:
claimName: ${NFS_NAME}
EOF
Below is an example how to point the other pods to this NFS. In particular, refer to the volumes section at the end of the YAML:
export NFS_NAME="nfs-share"
export NFS_IP=$(kubectl get --template={{.spec.clusterIP}} service/$NFS_NAME)
kubectl apply -f - <<EOF
kind: Deployment
apiVersion: apps/v1
metadata:
name: apache
labels:
app: apache
spec:
replicas: 2
selector:
matchLabels:
app: apache
template:
metadata:
labels:
app: apache
containers:
- name: apache
image: apache
ports:
- containerPort: 80
volumeMounts:
- mountPath: /var/www/html/
name: nfs-vol
subPath: html
volumes:
- name: nfs-vol
nfs:
server: $NFS_IP
path: /
EOF
It is correct that there is a 1:1 relation between a PersistentVolumeClaim and a PersistentVolume.
However, Pods running on the same Node can concurrently mount the same volume, e.g. use the same PersistentVolumeClaim.
If you use Minikube for local development, you only have one node, so you can use the same PersistentVolumeClaim. Since you want to use different files for each app, you could use a unique directory for each app in that shared volume.
So finally, I did it by using a dynamic provider.
I installed the stable/nfs-server-provisioner with helm. With proper configuration, it managed to create a pv named nfs two which my pvc's are able to bound :)
helm install stable/nfs-server-provisioner --name nfs-provisioner -f nfs_provisioner.yaml
the nfs_provisioner.yaml is as follows
persistence:
enabled: true
storageClass: "standard"
size: 20Gi
storageClass:
# Name of the storage class that will be managed by the provisioner
name: nfs
defaultClass: true
I am unable to create a mongo container using a deployment:
Persistent volume and persistent volume claims are working fine.
> kubectl logs -f mongo-depl-dc764fb6d-qqdxh
Error from server (BadRequest): container "mongo" in pod "mongo-depl-dc764fb6d-qqdxh" is waiting to start: CreateContainerError
Persistent Volume :
apiVersion: v1
kind: PersistentVolume
metadata:
name: mongo-pv
labels:
type: local
spec:
storageClassName: manual
capacity:
storage: 1Gi
accessModes:
- ReadWriteOnce
hostPath:
path: "E:\\Linux\\mongo"
Persistent Volume Claim :
apiVersion: v1
kind: PersistentVolumeClaim
metadata:
name: mongo-pvc
spec:
storageClassName: manual
accessModes:
- ReadWriteOnce
resources:
requests:
storage: 0.5Gi
Mongodb Deployment :
apiVersion: apps/v1
kind: Deployment
metadata:
name: mongo-depl
spec:
replicas: 1
selector:
matchLabels:
app: mongo
template:
metadata:
labels:
app: mongo
spec:
volumes:
- name: mongo-volume
persistentVolumeClaim:
claimName: mongo-pvc
containers:
- name: mongo
image: mongo:3.6.5-jessie
ports:
- name: mongo
containerPort: 27017
volumeMounts:
- name: mongo-volume
mountPath: /data/db
---
apiVersion: v1
kind: Service
metadata:
name: mongo-srv
spec:
type: NodePort
selector:
app: mongo
ports:
- name: mongo
protocol: TCP
port: 27017
targetPort: 27017
I would really appreciate the help. I am running kubernetes on a dev environment on the windows machine.
I am running local k8s cluster and defining PV as hostPath for mysql pods.
Sharing all the configuration details below .
kind: PersistentVolume
apiVersion: v1
metadata:
name: mysql-pv-volume
labels:
type: local
spec:
storageClassName: manual
capacity:
storage: 2Gi
accessModes:
- ReadWriteOnce
hostPath:
path: "/mnt/data"
---
apiVersion: v1
kind: PersistentVolumeClaim
metadata:
name: mysql-pv-claim
spec:
storageClassName: manual
accessModes:
- ReadWriteOnce
resources:
requests:
storage: 2Gi
---
apiVersion: v1
kind: Service
metadata:
name: mysql
spec:
ports:
- port: 3306
selector:
app: mysql
clusterIP: None
---
apiVersion: apps/v1 # for versions before 1.9.0 use apps/v1beta2
kind: Deployment
metadata:
name: mysql
spec:
selector:
matchLabels:
app: mysql
strategy:
type: Recreate
template:
metadata:
labels:
app: mysql
spec:
containers:
- image: mysql:5.7
name: mysql
env:
# Use secret in real usage
- name: MYSQL_ROOT_PASSWORD
value: password
ports:
- containerPort: 3306
name: mysql
volumeMounts:
- name: mysql-persistent-storage
mountPath: /var/lib/mysql
volumes:
- name: mysql-persistent-storage
persistentVolumeClaim:
claimName: mysql-pv-claim
The problem I am getting is as mysql pod is running in k8s cluster ,when its deleted and recreate ,it will choose any one of the node and deployed .So mysql hostpath always mounted to specific node .Is it a good idea to fix the node for mysql or any other options are there ?please share if any idea .
you have below choices
Use node selector or node affinity to ensure that pod gets scheduled on the node where the mount is created OR
Use local persistent volumes. it is supported on kubernetes 1.14 and above
Why are you using a PVC and a PV? Actually, for hostPath, you don't even need to create the PV object. It just gets it.
You should use a StatefulSet if you want a pod that is re-created to get the storage it was using the previous one (state).
apiVersion: apps/v1
kind: StatefulSet
metadata:
name: web
spec:
selector:
matchLabels:
app: mysql
serviceName: "mysql"
replicas: 1
template:
metadata:
labels:
app: mysql
spec:
terminationGracePeriodSeconds: 10
containers:
- name: mysql
image: mysql:5.7
env:
- name: MYSQL_ROOT_PASSWORD
value: password
ports:
- containerPort: 3306
name: mysql
volumeMounts:
- name: mysql-persistent-storage
mountPath: /var/lib/mysql
volumeClaimTemplates:
- metadata:
name: mysql-persistent-storage
spec:
accessModes: ["ReadWriteOnce"]
# storageClassName: "standard"
resources:
requests:
storage: 2Gi
This statefulSet fails, but it is a mysql thing. As reference, should serve.
I want to integrate a minio object storage in to my minikune cluster.
I use the docker file from the minio gitrepo
I also added the persistent volume with the claim
kind: PersistentVolume
apiVersion: v1
metadata:
name: minio-pv-volume
labels:
type: local
spec:
storageClassName: manual
capacity:
storage: 10Gi
volumeMode: Filesystem
accessModes:
- ReadWriteOnce
persistentVolumeReclaimPolicy: Retain
hostPath:
path: "/mnt/data/minio"
---
kind: PersistentVolumeClaim
apiVersion: v1
metadata:
name: minio-pv-claim
spec:
storageClassName: manual
accessModes:
- ReadWriteOnce
resources:
requests:
storage: 3Gi
for the minio deployment I have
apiVersion: apps/v1 # for versions before 1.9.0 use apps/v1beta2
kind: Deployment
metadata:
name: minio
spec:
selector:
matchLabels:
app: minio
role: master
tier: backend
replicas: 1
template:
metadata:
labels:
app: minio
role: master
tier: backend
spec:
imagePullSecrets:
- name: regcred
containers:
- name: minio
image: <secret Registry >
env:
- name: MINIO_ACCESS_KEY
value: akey
- name: MINIO_SECRET_KEY
value: skey
ports:
- containerPort: 9000
volumeMounts:
- name: data
mountPath: /data/ob
volumes:
- name: data
persistentVolumeClaim:
claimName: minio-pv-claim
For the service I opened up the external IP just for debugging
apiVersion: v1
kind: Service
metadata:
name: minio
labels:
app: minio
role: master
tier: backend
spec:
ports:
- port: 9000
targetPort: 9000
externalIPs:
- 192.168.99.101
selector:
app: minio
role: master
tier: backend
But when I start the deployment I get the error message ERROR Unable to initialize backend: The disk size is less than the minimum threshold.
I assumed that 3GB should be enough. How can I solve this issue moreover now that I try to delete my persistent volume it rest in the terminating status.
How can I run minio in a minikube clutster?
I dont think there is enough storage in /mnt/data inside minikube. Try /mnt/sda1 or /data. Better yet, go inside minikube and check the storage available. To get into minikube you can do minikube ssh.