Dynamic provisioning of PVCs in AKS

This document describes how to dynamically provision a volume using Kubernetes and Portworx.

Using Dynamic Provisioning

Using Dynamic Provisioning and Storage Classes you don't need to create Portworx volumes out of band and they will be created automatically. Using Storage Classes objects an admin can define the different classes of Portworx Volumes that are offered in a cluster. Following are the different parameters that can be used to define a Portworx Storage Class

Name	Description	Example
fs	Specifies a filesystem to be laid out: xfs|ext4. Note: The only preferred and default filesystem across all backends is ext4. The default backend can also support xfs. However, the PX-StoreV2 backend exclusively supports ext4, with no provision for any other filesystems.	fs: "ext4"
repl	Specifies the replication factor for the volume: 1|2|3	repl: "3"
sharedv4	Creates a globally shared namespace volume which can be used by multiple pods over NFS with POSIX compliant semantics	sharedv4: "true"
sharedv4_svc_type	Indicates the mechanism Kubernetes will use for locating your sharedv4 volume. If you use this flag and there's a failover of the nodes running your sharedv4 volume, you no longer need to restart your pods. Possible values are: ClusterIP or LoadBalancer.	sharedv4_svc_type: "ClusterIP"
sharedv4_failover_strategy	Specifies how aggressively to fail over to a new server for a Sharedv4 or Sharedv4 Service volume (Valid Values: aggressive, normal). The default failover strategy for sharedv4 service volumes is aggressive, because these volumes are able to fail over without restarting all the application pods. For more information, see Sharedv4 failover and failover strategy.	sharedv4_failover_strategy: "normal"
priority_io	Specifies IO Priority: low|medium|high. The default is low	priority_io: "high"
io_profile	Overrides I/O algorithm that Portworx uses for a volume. For more information about IO profiles, see the IO profiles section of the documentation.	io_profile: "db"
group	Specifies the group a volume should belong too. Portworx restricts replication sets of volumes of the same group on different nodes. If the force group option 'fg' is set to true, the volume group rule is strictly enforced. By default, it's not strictly enforced.	group: "volgroup1"
fg	Enforces volume group policy. If a volume belonging to a group cannot find nodes for its replication sets which don't have other volumes of the same group, the volume creation will fail.	fg: "true"
label	Arbitrary key: value labels that can be applied on a volume	label: "name: mypxvol"
nodes	Specifies comma-separated Portworx Node IDs to use for replication sets of the volume	nodes: "minion1,minion2"
ephemeral	Creates the ephemeral volumes	ephemeral: false
size	Specifies a volume size in GB (default 1)	size: "1073741824"
block_size	Specifies a block size in Bytes (default 4096)	block_size: "4096"
queue_depth	Specifies a block device queue depth. (Valid Range: [1 256]) (default 128)	queue_depth: 128
snap_interval	Specifies an interval in minutes at which periodic snapshots will be triggered. Set to 0 to disable snapshots
snap_schedule	Specifies the name of the snapshot schedule policy created using the pxctl sched-policy command	Refer to this page for examples
zones	Specify comma-separated zone names in which the volume replicas should be distributed
racks	Specify comma-separated rack names in which the volume replicas should be distributed
async_io	Enables asynchronous IO processing on the backend drives and could be useful in situations where the workload profile is bursty in nature. Please reach out to Portworx Support before enabling.	async_io: false
csi_mount_options	Specifies the mounting options for a volume through CSI
sharedv4_mount_options	Specifies a comma-separated list of Sharedv4 NFS client mount options provided as key=value pairs
proxy_endpoint	Specifies the endpoint address of the external NFS share Portworx is proxying	proxy_endpoint: "nfs://<nfs-share-endpoint>"
proxy_nfs_subpath	Specifies the sub-path from the NFS share to which this proxy volume has access to
proxy_nfs_exportpath	Exports path for NFS proxy volume	proxy_nfs_exportpath: "/<mount-path>"
export_options	Defines the export options. Currently, only NFS export options are supported for Sharedv4 volumes
mount_options	Specifies the mounting options for a volume when it is attached and mounted
best_effort_location_provisioning	Requested nodes, zones, racks are optional
direct_io	Enables Direct IO on a volume	direct_io: "true"
scan_policy_trigger	Specifies the trigger point on which filesystem check is triggered. Valid Values: none, on_mount, on_next_mount
scan_policy_action	Specifies a filesystem scan action to be taken when triggered. Valid Values: none, scan_only, scan_repair
force_unsupported_fs_type	Forces a filesystem type that is not supported. The driver may still refuse to use the type	force_unsupported_fs_type: false
match_src_vol_provision	Provisions the restore volume on the same pools as the source volume (src volume must exist)
nodiscard	Mounts the volume with nodiscard option. This is useful when the volume undergoes a large amount of block discards and later the application rewrites to these discarded block making the discard work done by Portworx useless. This option must be used along with auto_fstrim. Refer to this page for limitations on xfs formatted volumes.	nodiscard: false
auto_fstrim	Enables auto_fstrim on a volume and requires the nodiscard option to be set. Refer to this page for more details.	auto_fstrim: true
storagepolicy	Creates a volume on the Portworx cluster that follows the specified set of specs/rules. Refer this page for more details.
backend	Specifies which storage backend Portworx is going to provide direct access to. (Valid Values: pure_block, pure_file)	backend: "pure_block"
pure_export_rules	Specifies the export rules for exporting a Pure Flashblade volume	pure_export_rules: "*(rw)"
io_throttle_rd_iops	Specifies maximum Read IOPs a volume will be throttled to. Refer to this page for more details.	io_throttle_rd_iops: "1024"
io_throttle_wr_iops	Specifies maximum Write IOPs a volume will be throttled to. Refer to this page for more details.	io_throttle_wr_iops: "1024"
io_throttle_rd_bw	Specifies maximum Read bandwidth a volume will be throttled to. Refer to this page for more details.	io_throttle_rd_bw: "10"
io_throttle_wr_bw	Specifies maximum Write bandwidth a volume will be throttled to. Refer to this page for more details.	io_throttle_wr_bw: "10"
aggregation_level	Specifies the number of replication sets the volume can be aggregated from	aggregation_level: "2"
sticky	Creates sticky volumes that cannot be deleted until the flag is disabled	sticky: "true"
journal	Indicates if you want to use journal device for the volume's data. This will use the journal device that you used when installing Portworx. This is useful to absorb frequent syncs from short bursty workloads. Default: false	journal: "true"
secure	Creates an encrypted volume. For details about how you can create encrypted volumes, see the Create encrypted PVCs page.	secure: "true"
placement_strategy	Flag to refer the name of the VolumePlacementStrategy. For example: apiVersion: storage.k8s.io/v1 kind: StorageClass metadata:  name: postgres-storage-class provisioner: pxd.portworx.com parameters:  placement_strategy: "postgres-volume-affinity" For details about how to create and use VolumePlacementStrategy, see [this page] (/docs/portworx-enterprise/operations/operate-kubernetes/storage-operations/create-pvcs/volume-placement-strategies/create-use-volplacestrat.md).	placement_strategy: "postgres-volume-affinity"
snapshotschedule.stork.libopenstorage.org	Creates scheduled snapshots with Stork. For example: snapshotschedule.stork.libopenstorage.org/default-schedule:   schedulePolicyName: daily   annotations:     portworx/snapshot-type: local snapshotschedule.stork.libopenstorage.org weekly-schedule:   schedulePolicyName: weekly   annotations:     portworx/snapshot-type: cloud     portworx/cloud-cred-id: <credential-uuid> NOTE:This example references two schedules: The default-schedule backs up volumes to the local Portworx cluster daily. The weekly-schedule backs up volumes to cloud storage every week. For details about how you can create scheduled snapshots with Stork, see the Scheduled snapshots page.

note

For the list of Kubernetes-specific parameters that you can use with a Portworx Storage class, see the Storage Classes page of the Kubernetes documentation.

Provision volumes

Step 1: Create Storage Class.

Create the storageclass:

kubectl create -f portworx-sc.yaml

Example:

kind: StorageClass
apiVersion: storage.k8s.io/v1
metadata:
  name: portworx-sc
provisioner: pxd.portworx.com
parameters:
  repl: "1"

Download example

Verifying storage class is created:

kubectl describe storageclass portworx-sc

     Name: 	        	portworx-sc
     IsDefaultClass:	No
     Annotations:		<none>
     Provisioner:		pxd.portworx.com
     Parameters:		repl=1
     No events.

Step 2: Create Persistent Volume Claim.

Creating the persistent volume claim:

kubectl create -f portworx-volume-pvcsc.yaml

Example:

kind: PersistentVolumeClaim
apiVersion: v1
metadata:
  name: pvcsc001
  annotations:
    volume.beta.kubernetes.io/storage-class: portworx-sc
spec:
  accessModes:
    - ReadWriteOnce
  resources:
    requests:
      storage: 2Gi

Download example

Verifying persistent volume claim is created:

kubectl describe pvc pvcsc001

Name:	      	pvcsc001
Namespace:      default
StorageClass:   portworx-sc
Status:	      	Bound
Volume:         pvc-xxxxxxxx-xxxx-xxxx-xxxx-08002729a32b
Labels:	      	<none>
Capacity:	      2Gi
Access Modes:     RWO
No Events.

Persistent Volume is automatically created and is bounded to this pvc.

Verifying persistent volume is created:

kubectl describe pv pvc-xxxxxxxx-xxxx-xxxx-xxxx-08002729a32b

Name: 	      	pvc-xxxxxxxx-xxxx-xxxx-xxxx-08002729a32b
Labels:        	<none>
StorageClass:  	portworx-sc
Status:	      	Bound
Claim:	      	default/pvcsc001
Reclaim Policy: Delete
Access Modes:     RWO
Capacity:	     2Gi
Message:
Source:
Type:	      	PortworxVolume (a Portworx Persistent Volume resource)
VolumeID:   	374093969022973811
No events.

Step 3: Create Pod which uses Persistent Volume Claim with storage class.

Create the pod:

kubectl create -f portworx-volume-pvcscpod.yaml

Example:

apiVersion: v1
kind: Pod
metadata:
  name: pvpod
spec:
  containers:
    - name: test-container
      image: gcr.io/google_containers/test-webserver
      volumeMounts:
        - name: test-volume
          mountPath: /test-portworx-volume
  volumes:
    - name: test-volume
      persistentVolumeClaim:
        claimName: pvcsc001

Download example

Verifying pod is created:

kubectl get pod pvpod

NAME      READY     STATUS    RESTARTS   AGE
pvpod       1/1     Running   0          48m

note

To access PV/PVCs with a non-root user refer here

Delete volumes

For dynamically provisioned volumes using StorageClass and PVC (PersistenVolumeClaim), if a PVC is deleted, the corresponding Portworx volume will also get deleted. This is because Kubernetes, for PVC, creates volumes with a reclaim policy of deletion. So the volumes get deleted on PVC deletion.

To delete the PVC and the volume, you can run kubectl delete -f <pvc_spec_file.yaml>