Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Add notes about use of writable snapshots for PowerScale #1453

Closed
wants to merge 4 commits into from
Closed

Add notes about use of writable snapshots for PowerScale #1453

Changes from all commits
Commits
Show all changes
4 commits
Select commit Hold shift + click to select a range
88ac9ee
Add notes about use of writable snapshots.
donatwork Feb 17, 2025
a408a88
Address rsedlock's comments.
donatwork Feb 18, 2025
4ae8c96
Address grammar item.
donatwork Feb 18, 2025
a99f2d2
Fix grammar.
donatwork Feb 18, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
38 changes: 23 additions & 15 deletions content/docs/csidriver/features/powerscale.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -66,7 +66,7 @@ spec:
requests:
storage: 5Gi
volumeName: isilonstaticpv
storageClassName: isilon
storageClassName: isilon
```

4. Then use this PVC as a volume in a pod.
Expand Down Expand Up @@ -108,7 +108,7 @@ spec:
resources:
requests:
storage: 5Gi
storageClassName: isilon
storageClassName: isilon
```

## Volume Snapshot Feature
Expand Down Expand Up @@ -150,7 +150,7 @@ parameters:

The following is a sample manifest for creating a Volume Snapshot using the **v1** snapshot APIs; The following snippet assumes that the persistent volume claim name is testvolume.

```yaml
```yaml
apiVersion: snapshot.storage.k8s.io/v1
kind: VolumeSnapshot
metadata:
Expand Down Expand Up @@ -198,6 +198,10 @@ spec:

> Starting from CSI PowerScale driver version 2.2, different isi paths can be used to create PersistentVolumeClaim from VolumeSnapshot.This means the isi paths of the new volume and the VolumeSnapshot can be different.

>**Note:** When the access mode of ReadWriteMany is used, a writable snapshot is created using the read only snapshot as the source. Creation of snapshots from PVCs which are created from snapshots is not supported.

>**Note:** Older releases of OneFS limit the number of writable snapshots to a maximum of thirty per cluster. Newer releases of OneFS increased the limit to 1024 writable snapshots per cluster. Consult your Dell PowerStore Technical representative if you need to increase the number of writable snapshots.

## Volume Expansion

CSI PowerScale driver version 1.2 and later supports the expansion of Persistent Volumes (PVs). This expansion can be done either online (for example, when a PVC is attached to a node) or offline (for example, when a PVC is not attached to any node).
Expand Down Expand Up @@ -287,6 +291,10 @@ spec:
apiGroup: ""
```

During the volume cloning process a read only snapshot of the source volume is made and a writable snapshot is created using the read only snapshot as the source. Snapshots provide way to quickly create copies of source volumes. A SnapshotIQ license is required for snapshots to be created. Note that cloning of clones is not currently supported.

>**Note:** Older releases of OneFS limit the number of writable snapshots to a maximum of thirty per cluster. Newer releases of OneFS increased the limit to 1024 writable snapshots per cluster. Consult your Dell PowerStore Technical representative if you need to increase the number of writable snapshots.

## Controller HA

CSI PowerScale driver version 1.4.0 and later supports running multiple replicas of the controller pod. Leader election is only applicable for all sidecar containers and driver container will be running in all controller pods. In case of a failure, one of the standby pods becomes active and takes the position of leader. This is achieved by using native leader election mechanisms utilizing `kubernetes leases`.
Expand All @@ -301,10 +309,10 @@ controllerCount: 2

>**NOTE:** The default value for controllerCount is 2. It is recommended to not change this unless really required. Also, if the controller count is greater than the number of available nodes (where the pods can be scheduled), some controller pods will remain in a Pending state.

If you are using the Dell CSM Operator, the value to adjust is:
If you are using the Dell CSM Operator, the value to adjust is:

```yaml
replicas: 2
replicas: 2
```

For more details about configuring Controller HA using the Dell CSM Operator, see the [Dell CSM Operator documentation](../../../deployment/csmoperator/#custom-resource-specification).
Expand Down Expand Up @@ -374,12 +382,12 @@ To utilize the Topology feature, create a custom `StorageClass` with `volumeBind

```yaml
# This is a sample manifest for utilizing the topology feature and mount options.
# PVCs created using this storage class will be scheduled
# PVCs created using this storage class will be scheduled
# only on the nodes with access to Isilon

# Change all instances of <ISILON_IP> to the IP of the PowerScale OneFS API server

# Provide mount options through "mountOptions" attribute
# Provide mount options through "mountOptions" attribute
# to create PVCs with mount options.

apiVersion: storage.k8s.io/v1
Expand All @@ -394,9 +402,9 @@ parameters:
IsiPath: "/ifs/data/csi"
# AccessZone groupnet service IP. Update AzServiceIP in values.yaml if different than isiIP.
#AzServiceIP : 192.168.2.1
# When a PVC is being created, it takes the storage class' value of "storageclass.rootClientEnabled",
# which determines, when a node mounts the PVC, in NodeStageVolume, whether to add the k8s node to
# the "Root clients" field (when true) or "Clients" field (when false) of the NFS export
# When a PVC is being created, it takes the storage class' value of "storageclass.rootClientEnabled"
# which determines (when a node mounts the PVC, in NodeStageVolume) whether to add the k8s node to
# the "Root clients" field (when true) or "Clients" field (when false) of the NFS export
RootClientEnabled: "false"
# Name of PowerScale cluster where pv will be provisioned
# This name should match with name of one of the cluster configs in isilon-creds secret
Expand Down Expand Up @@ -453,7 +461,7 @@ CSI for PowerScale driver version 2.8.0 and above supports Storage Capacity Trac
This feature helps the scheduler to make more informed choices about where to schedule pods which depends on unbound volumes with late binding (aka "wait for first consumer"). Pods will be scheduled on a node (satisfying the topology constraints) only if the requested capacity is available on the storage array.
If such a node is not available, the pods stay in Pending state. This means pods are not scheduled.

Without storage capacity tracking, pods get scheduled on a node satisfying the topology constraints. If the required capacity is not available, volume attachment to the pods fails, and pods remain in ContainerCreating state. Storage capacity tracking eliminates unnecessary scheduling of pods when there is insufficient capacity.
Without storage capacity tracking, pods get scheduled on a node satisfying the topology constraints. If the required capacity is not available, volume attachment to the pods fails, and pods remain in ContainerCreating state. Storage capacity tracking eliminates unnecessary scheduling of pods when there is insufficient capacity.

The attribute `storageCapacity.enabled` in `values.yaml` can be used to enable/disable the feature during driver installation using helm. This is by default set to true. To configure how often driver checks for changed capacity set `storageCapacity.pollInterval` attribute. In case of driver installed via operator, this interval can be configured in the sample file provided [here.](https://github.com/dell/csm-operator/blob/main/samples/) by editing the `--capacity-poll-interval` argument present in the provisioner sidecar.

Expand Down Expand Up @@ -519,12 +527,12 @@ parameters:
#Optional: true
#Default value: Limit not Set
#AdvisoryLimit: "50"

#Parameter to set soft limit to quota
#Optional: true
#Default value: Limit not Set
#SoftLimit: "80"

#Parameter which must be mentioned along with Soft Limit
#Soft Limit can be exceeded until the grace period
#Optional: true
Expand Down Expand Up @@ -609,7 +617,7 @@ To update the log level dynamically user has to edit the ConfigMap `isilon-confi

```bash
kubectl edit configmap -n isilon isilon-config-params
```
```

>Note: Prior to CSI Driver for PowerScale version 2.0.0, the log level was allowed to be updated dynamically through `logLevel` attribute in the secret object.

Expand Down Expand Up @@ -661,7 +669,7 @@ For example, if a volume were to be deleted from the array, or unmounted outside
2. For controller plugin, by setting attribute `controller.healthMonitor.enabled` to `true` in `values.yaml` file. Also health monitoring interval can be changed through attribute `controller.healthMonitor.interval` in `values.yaml` file.
3. For node plugin, by setting attribute `node.healthMonitor.enabled` to `true` in `values.yaml` file.

## Single Pod Access Mode for PersistentVolumes- ReadWriteOncePod
## Single Pod Access Mode for PersistentVolumes- ReadWriteOncePod

Use `ReadWriteOncePod(RWOP)` access mode if you want to ensure that only one pod across the whole cluster can read that PVC or write to it. This is supported for CSI Driver for PowerScale 2.1.0+ and Kubernetes version 1.22+.

Expand Down
Loading