Does local-path super multiple sc for deffrence performance disk

[zhfg]

I have hdd and ssd installed in host server.
I want create two k8s storageclass to support for both hdd and ssd to adapt different performance requirement.
Is local-path supported this and how to do with it

[wojtasool]

I would say that you can create 2 or more installations of Local Path Provisioners each with different disks configured and with different names for each Storage Class. This way you can reference different disk types using specific SC.

[zhfg]

Thanks for your replay,
I can not found any documents about what you told me. for example, how to define provisioner myself.
I will do more research. and left a update i get any move forward

[qianyidui]

apiVersion: v1
kind: Namespace
metadata:
name: local-path-storage-ssd

---

apiVersion: v1
kind: ServiceAccount
metadata:
name: local-path-provisioner-service-account
namespace: local-path-storage-ssd

---

apiVersion: rbac.authorization.k8s.io/v1
kind: ClusterRole
metadata:
name: local-path-provisioner-role-ssd
rules:

• apiGroups: [ "" ]
  resources: [ "nodes", "persistentvolumeclaims", "configmaps" ]
  verbs: [ "get", "list", "watch" ]
• apiGroups: [ "" ]
  resources: [ "endpoints", "persistentvolumes", "pods" ]
  verbs: [ "*" ]
• apiGroups: [ "" ]
  resources: [ "events" ]
  verbs: [ "create", "patch" ]
• apiGroups: [ "storage.k8s.io" ]
  resources: [ "storageclasses" ]
  verbs: [ "get", "list", "watch" ]

---

apiVersion: rbac.authorization.k8s.io/v1
kind: ClusterRoleBinding
metadata:
name: local-path-provisioner-bind-ssd
roleRef:
apiGroup: rbac.authorization.k8s.io
kind: ClusterRole
name: local-path-provisioner-role-ssd
subjects:

• kind: ServiceAccount
  name: local-path-provisioner-service-account
  namespace: local-path-storage-ssd

---

apiVersion: apps/v1
kind: Deployment
metadata:
name: local-path-provisioner-ssd
namespace: local-path-storage-ssd
spec:
replicas: 1
selector:
matchLabels:
app: local-path-provisioner
template:
metadata:
labels:
app: local-path-provisioner
spec:
serviceAccountName: local-path-provisioner-service-account
containers:
- name: local-path-provisioner
image: rancher/local-path-provisioner:v0.0.22
imagePullPolicy: IfNotPresent
command:
- local-path-provisioner
- --debug
- start
- --config
- /etc/config/config.json
volumeMounts:
- name: config-volume
mountPath: /etc/config/
env:
- name: POD_NAMESPACE
valueFrom:
fieldRef:
fieldPath: metadata.namespace
- name: PROVISIONER_NAME
value: rancher.io/local-path-ssd
volumes:
- name: config-volume
configMap:
name: local-path-config

---

apiVersion: storage.k8s.io/v1
kind: StorageClass
metadata:
name: local-path-ssd
provisioner: rancher.io/local-path-ssd
volumeBindingMode: WaitForFirstConsumer
reclaimPolicy: Delete

---

kind: ConfigMap
apiVersion: v1
metadata:
name: local-path-config
namespace: local-path-storage-ssd
data:
config.json: |-
{
"nodePathMap":[
{
"node":"k3s",
"paths":["/raid1"]
}
]
}
setup: |-
#!/bin/sh
set -eu
mkdir -m 0777 -p "$VOL_DIR"
teardown: |-
#!/bin/sh
set -eu
rm -rf "$VOL_DIR"
helperPod.yaml: |-
apiVersion: v1
kind: Pod
metadata:
name: helper-pod
spec:
containers:
- name: helper-pod
image: busybox
imagePullPolicy: IfNotPresent

[qianyidui]

apiVersion: v1
kind: Namespace
metadata:
name: local-path-storage-ssd

---

apiVersion: v1
kind: ServiceAccount
metadata:
name: local-path-provisioner-service-account
namespace: local-path-storage-ssd

---

apiVersion: rbac.authorization.k8s.io/v1
kind: ClusterRole
metadata:
name: local-path-provisioner-role-ssd
rules:

• apiGroups: [ "" ]
  resources: [ "nodes", "persistentvolumeclaims", "configmaps" ]
  verbs: [ "get", "list", "watch" ]
• apiGroups: [ "" ]
  resources: [ "endpoints", "persistentvolumes", "pods" ]
  verbs: [ "*" ]
• apiGroups: [ "" ]
  resources: [ "events" ]
  verbs: [ "create", "patch" ]
• apiGroups: [ "storage.k8s.io" ]
  resources: [ "storageclasses" ]
  verbs: [ "get", "list", "watch" ]

---

apiVersion: rbac.authorization.k8s.io/v1
kind: ClusterRoleBinding
metadata:
name: local-path-provisioner-bind-ssd
roleRef:
apiGroup: rbac.authorization.k8s.io
kind: ClusterRole
name: local-path-provisioner-role-ssd
subjects:

• kind: ServiceAccount
  name: local-path-provisioner-service-account
  namespace: local-path-storage-ssd

---

apiVersion: apps/v1
kind: Deployment
metadata:
name: local-path-provisioner-ssd
namespace: local-path-storage-ssd
spec:
replicas: 1
selector:
matchLabels:
app: local-path-provisioner
template:
metadata:
labels:
app: local-path-provisioner
spec:
serviceAccountName: local-path-provisioner-service-account
containers:
- name: local-path-provisioner
image: rancher/local-path-provisioner:v0.0.22
imagePullPolicy: IfNotPresent
command:
- local-path-provisioner
- --debug
- start
- --config
- /etc/config/config.json
volumeMounts:
- name: config-volume
mountPath: /etc/config/
env:
- name: POD_NAMESPACE
valueFrom:
fieldRef:
fieldPath: metadata.namespace
- name: PROVISIONER_NAME
value: rancher.io/local-path-ssd
volumes:
- name: config-volume
configMap:
name: local-path-config

---

apiVersion: storage.k8s.io/v1
kind: StorageClass
metadata:
name: local-path-ssd
provisioner: rancher.io/local-path-ssd
volumeBindingMode: WaitForFirstConsumer
reclaimPolicy: Delete

---

kind: ConfigMap
apiVersion: v1
metadata:
name: local-path-config
namespace: local-path-storage-ssd
data:
config.json: |-
{
"nodePathMap":[
{
"node":"k3s",
"paths":["/raid1"]
}
]
}
setup: |-
#!/bin/sh
set -eu
mkdir -m 0777 -p "$VOL_DIR"
teardown: |-
#!/bin/sh
set -eu
rm -rf "$VOL_DIR"
helperPod.yaml: |-
apiVersion: v1
kind: Pod
metadata:
name: helper-pod
spec:
containers:
- name: helper-pod
image: busybox
imagePullPolicy: IfNotPresent

[qianyidui]

env:

• name: POD_NAMESPACE
  valueFrom:
  fieldRef:
  fieldPath: metadata.namespace
• name: PROVISIONER_NAME
  value: rancher.io/local-path-ssd

[lholota]

I'm currently hitting the same thing, tbh this feels like it's semantically wrong. The point of a storage class is exactly to define classes of storage (i.e. fast=ssd, slower=hdd etc.) which may differ in parameters/options.

I believe the provisioner should take the paths from the storage class definition rather than from the centralized config map. This way the paths/sharedFileSystemPath could be defined per storageClass and you don't need to have multiple provisioner pods running inside of the same cluster (in my case it's 5 installations of the provisioner which seems rather cumbersome...)

A way to implement this in backward compatible way would be to let the user define these fields in both places and letting the storageClass behave as an override of the default values found in the configMap.

[lholota]

Hi all, any chance the suggestion above would be considered or is it too much of a blast into current design?

[LinkMaq]

I think the local-path-provisioner was designed to face simple scenarios from the beginning. If in a kubernetes cluster there are multiple performance disks provided through storageClass, using openebs/local-hostpath should be more reasonable.

[ChristianCiach]

This should be fixed by #306, I think.

[laurivosandi]

The setup and teardown scripts, helper pod image can also differ per storage class thus #306 is short-sighted solution. I would recommend giving up the ConfigMap as it is at the moment and move the scrpits into StorageClass definition

[github-actions]

This issue is stale because it has been open 30 days with no activity. Remove stale label or comment or this will be closed in 5 days.