Local Path Provisioner provides a way for the Kubernetes users to utilize the local storage in each node. Based on the user configuration, the Local Path Provisioner will create either hostPath or local based persistent volume on the node automatically. It utilizes the features introduced by Kubernetes Local Persistent Volume feature, but makes it a simpler solution than the built-in local volume feature in Kubernetes.
Dynamic provisioning the volume using hostPath or local.
- Currently the Kubernetes Local Volume provisioner cannot do dynamic provisioning for the local volumes.
- Local based persistent volumes are an experimental feature (example usage).
- No support for the volume capacity limit currently.
- The capacity limit will be ignored for now.
Kubernetes v1.12+.
In this setup, the directory /opt/local-path-provisioner will be used across all the nodes as the path for provisioning (a.k.a, store the persistent volume data). The provisioner will be installed in local-path-storage namespace by default.
- Stable
kubectl apply -f https://raw.githubusercontent.com/rancher/local-path-provisioner/v0.0.30/deploy/local-path-storage.yaml
- Development
kubectl apply -f https://raw.githubusercontent.com/rancher/local-path-provisioner/master/deploy/local-path-storage.yaml
Or, use kustomize to deploy.
- Stable
kustomize build "github.com/rancher/local-path-provisioner/deploy?ref=v0.0.30" | kubectl apply -f -
- Development
kustomize build "github.com/rancher/local-path-provisioner/deploy?ref=master" | kubectl apply -f -
After installation, you should see something like the following:
$ kubectl -n local-path-storage get pod
NAME READY STATUS RESTARTS AGE
local-path-provisioner-d744ccf98-xfcbk 1/1 Running 0 7m
Check and follow the provisioner log using:
kubectl -n local-path-storage logs -f -l app=local-path-provisioner
Create a hostPath backend Persistent Volume and a pod uses it:
kubectl create -f https://raw.githubusercontent.com/rancher/local-path-provisioner/master/examples/pvc/pvc.yaml
kubectl create -f https://raw.githubusercontent.com/rancher/local-path-provisioner/master/examples/pod/pod.yaml
Or, use kustomize to deploy them.
kustomize build "github.com/rancher/local-path-provisioner/examples/pod?ref=master" | kubectl apply -f -
You should see the PV has been created:
$ kubectl get pv
NAME CAPACITY ACCESS MODES RECLAIM POLICY STATUS CLAIM STORAGECLASS REASON AGE
pvc-bc3117d9-c6d3-11e8-b36d-7a42907dda78 2Gi RWO Delete Bound default/local-path-pvc local-path 4s
The PVC has been bound:
$ kubectl get pvc
NAME STATUS VOLUME CAPACITY ACCESS MODES STORAGECLASS AGE
local-path-pvc Bound pvc-bc3117d9-c6d3-11e8-b36d-7a42907dda78 2Gi RWO local-path 16s
And the Pod started running:
$ kubectl get pod
NAME READY STATUS RESTARTS AGE
volume-test 1/1 Running 0 3s
Write something into the pod
kubectl exec volume-test -- sh -c "echo local-path-test > /data/test"
Now delete the pod using
kubectl delete -f https://raw.githubusercontent.com/rancher/local-path-provisioner/master/examples/pod/pod.yaml
After confirm that the pod is gone, recreated the pod using
kubectl create -f https://raw.githubusercontent.com/rancher/local-path-provisioner/master/examples/pod/pod.yaml
Check the volume content:
$ kubectl exec volume-test -- sh -c "cat /data/test"
local-path-test
Delete the pod and pvc
kubectl delete -f https://raw.githubusercontent.com/rancher/local-path-provisioner/master/examples/pod/pod.yaml
kubectl delete -f https://raw.githubusercontent.com/rancher/local-path-provisioner/master/examples/pvc/pvc.yaml
Or, use kustomize to delete them.
kustomize build "github.com/rancher/local-path-provisioner/examples/pod?ref=master" | kubectl delete -f -
The volume content stored on the node will be automatically cleaned up. You can check the log of local-path-provisioner-xxx for details.
Now you've verified that the provisioner works as expected.
The configuration of the provisioner is a json file config.json, a Pod template helperPod.yaml and two bash scripts setup and teardown, stored in a config map, e.g.:
kind: ConfigMap
apiVersion: v1
metadata:
name: local-path-config
namespace: local-path-storage
data:
config.json: |-
{
"nodePathMap":[
{
"node":"DEFAULT_PATH_FOR_NON_LISTED_NODES",
"paths":["/opt/local-path-provisioner"]
},
{
"node":"yasker-lp-dev1",
"paths":["/opt/local-path-provisioner", "/data1"]
},
{
"node":"yasker-lp-dev3",
"paths":[]
}
]
}
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:
priorityClassName: system-node-critical
tolerations:
- key: node.kubernetes.io/disk-pressure
operator: Exists
effect: NoSchedule
containers:
- name: helper-pod
image: busybox
The helperPod is allowed to run on nodes experiencing disk pressure conditions, despite the potential resource constraints. When it runs on such a node, it can carry out specific cleanup tasks, freeing up space in PVCs, and resolving the disk-pressure issue.
nodePathMap is the place user can customize where to store the data on each node.
- If one node is not listed on the
nodePathMap, and Kubernetes wants to create volume on it, the paths specified inDEFAULT_PATH_FOR_NON_LISTED_NODESwill be used for provisioning. - If one node is listed on the
nodePathMap, the specified paths inpathswill be used for provisioning.- If one node is listed but with
pathsset to[], the provisioner will refuse to provision on this node. - If more than one path was specified, the path would be chosen randomly when provisioning.
- If one node is listed but with
sharedFileSystemPath allows the provisioner to use a filesystem that is mounted on all nodes at the same time.
In this case all access modes are supported: ReadWriteOnce, ReadOnlyMany and ReadWriteMany for storage claims.
storageClassConfigs is a map from storage class names to objects containing nodePathMap or sharedFilesystemPath, as described above.
In addition volumeBindingMode: Immediate can be used in StorageClass definition.
Please note that nodePathMap, sharedFileSystemPath, and storageClassConfigs are mutually exclusive. If sharedFileSystemPath or stroageClassConfigs are used, then nodePathMap must be set to [].
The setupCommand and teardownCommand allow you to specify the path to binary files in helperPod that will be called when creating or deleting pvc respectively. This can be useful if you need to use distroless images for security reasons. See the examples/distroless directory for an example. A binary file can take the following parameters:
| Parameter | Description |
|---|---|
| -p | Volume directory that should be created or removed. |
| -m | The PersistentVolume mode (Block or Filesystem). |
| -s | Requested volume size in bytes. |
| -a | Action type. Can be create or delete |
The setupCommand and teardownCommand have higher priority than the setup and teardown scripts from the ConfigMap.
The configuration must obey following rules:
config.jsonmust be a valid json file.- A path must start with
/, a.k.a an absolute path. - Root directory(
/) is prohibited. - No duplicate paths allowed for one node.
- No duplicate node allowed.
- The
setupscript is run before the volume is created, to prepare the volume directory on the node. - The
teardownscript is run after the volume is deleted, to cleanup the volume directory on the node. - The
helperPod.yamltemplate is used to create a helper Pod that runs thesetuporteardownscript.
The scripts receive their input as environment variables:
| Environment variable | Description |
|---|---|
VOL_DIR |
Volume directory that should be created or removed. |
VOL_MODE |
The PersistentVolume mode (Block or Filesystem). |
VOL_SIZE_BYTES |
Requested volume size in bytes. |
The provisioner supports automatic configuration reloading. Users can change the configuration using kubectl apply or kubectl edit with config map local-path-config. There is a delay between when the user updates the config map and the provisioner picking it up. In order for this to occur for updates made to the helper pod manifest, the following environment variable must be added to the provisioner container. If not, then the manifest used for the helper pod will be the same as what was in the config map when the provisioner was last restarted/deployed.
- name: CONFIG_MOUNT_PATH
value: /etc/config/When the provisioner detects the configuration changes, it will try to load the new configuration. Users can observe it in the log
time="2018-10-03T05:56:13Z" level=debug msg="Applied config: {"nodePathMap":[{"node":"DEFAULT_PATH_FOR_NON_LISTED_NODES","paths":["/opt/local-path-provisioner"]},{"node":"yasker-lp-dev1","paths":["/opt","/data1"]},{"node":"yasker-lp-dev3"}]}"
If the reload fails, the provisioner will log the error and continue using the last valid configuration for provisioning in the meantime.
time="2018-10-03T05:19:25Z" level=error msg="failed to load the new config file: fail to load config file /etc/config/config.json: invalid character '#' looking for beginning of object key string"
time="2018-10-03T05:20:10Z" level=error msg="failed to load the new config file: config canonicalization failed: path must start with / for path opt on node yasker-lp-dev1"
time="2018-10-03T05:23:35Z" level=error msg="failed to load the new config file: config canonicalization failed: duplicate path /data1 on node yasker-lp-dev1
time="2018-10-03T06:39:28Z" level=error msg="failed to load the new config file: config canonicalization failed: duplicate node yasker-lp-dev3"
To specify the type of volume you want the provisioner to create, add either of the following annotations;
- PVC:
annotations:
volumeType: <local or hostPath>- StorageClass:
annotations:
defaultVolumeType: <local or hostPath>A few things to note; the annotation for the StorageClass will apply to all volumes using it and is superseded by the annotation on the PVC if one is provided. If neither of the annotations was provided then we default to hostPath.
If more than one paths are specified in the nodePathMap the path is chosen randomly. To make the provisioner choose a specific path, use a storageClass defined with a parameter called nodePath. Note that this path should be defined in the nodePathMap.
By default the volume subdirectory is named using the template {{ .PVName }}_{{ .PVC.Namespace }}_{{ .PVC.Name }} which make the directory specific to the PV instance. The template can be changed using the pathPattern parameter which is interpreted as a go template. The template has access to the PV name using the PVName variable and the PVC metadata object, including labels and annotations, with the PVC variable.
apiVersion: storage.k8s.io/v1
kind: StorageClass
metadata:
name: ssd-local-path
provisioner: rancher.io/local-path
parameters:
nodePath: /data/ssd
pathPattern: "{{ .PVC.Namespace }}/{{ .PVC.Name }}"
volumeBindingMode: WaitForFirstConsumer
reclaimPolicy: Delete
Here the provisioner will use the path /data/ssd with a subdirectory per namespace and PVC when storage class ssd-local-path is used.
Before uninstallation, make sure the PVs created by the provisioner have already been deleted. Use kubectl get pv and make sure no PV with StorageClass local-path.
To uninstall, execute:
- Stable
kubectl delete -f https://raw.githubusercontent.com/rancher/local-path-provisioner/v0.0.30/deploy/local-path-storage.yaml
- Development
kubectl delete -f https://raw.githubusercontent.com/rancher/local-path-provisioner/master/deploy/local-path-storage.yaml
it providers a out-of-cluster debug env for developers
git clone https://github.com/rancher/local-path-provisioner.git
cd local-path-provisioner
go build
kubectl apply -f debug/config.yaml
./local-path-provisioner --debug start --service-account-name=defaultkubectl delete -f debug/config.yaml
Copyright (c) 2014-2020 Rancher Labs, Inc.
Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except in compliance with the License. You may obtain a copy of the License at
http://www.apache.org/licenses/LICENSE-2.0
Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the specific language governing permissions and limitations under the License.
