Skip to main content
A newer release of this product is available.

Storage configuration options

Contributors juliantap netapp-mwallis
PDFs

See the configuration options available for your Astra Trident configurations.

Global configuration options

These configuration options apply to all Astra Trident configurations, regardless of the storage platform being used.

Option Description Example

version

Config file version number

1

storageDriverName

Name of storage driver

ontap-nas, ontap-san, ontap-nas-economy,
ontap-nas-flexgroup, solidfire-san, azure-netapp-files, or gcp-cvs

storagePrefix

Optional prefix for volume names. Default: “netappdvp_”.

staging_

limitVolumeSize

Optional restriction on volume sizes. Default: “” (not enforced)

10g
Tip Do not use storagePrefix (including the default) for Element backends. By default, the solidfire-san driver will ignore this setting and not use a prefix. We recommend using either a specific tenantID for Docker volume mapping or using the attribute data which is populated with the Docker version, driver info, and raw name from Docker in cases where any name munging may have been used.

Default options are available to avoid having to specify them on every volume you create. The size option is available for all the controller types. See the ONTAP configuration section for an example of how to set the default volume size.

Option Description Example

size

Optional default size for new volumes. Default: “1G”

10G

ONTAP configuration

In addition to the global configuration values above, when using ONTAP, the following top-level options are available.

Option Description Example

managementLIF

IP address of ONTAP management LIF. You can specify a fully-qualified domain name (FQDN).

10.0.0.1

dataLIF

IP address of protocol LIF; will be derived if not specified. For the ontap-nas drivers only, you can specify an FQDN, in which case the FQDN will be used for the NFS mount operations. For the ontap-san drivers, the default is to use all data LIF IPs from the SVM and to use iSCSI multipath. Specifying an IP address for dataLIF for the ontap-san drivers forces the driver to disable multipath and use only the specified address.

10.0.0.2

svm

Storage virtual machine to use (required, if management LIF is a cluster LIF)

svm_nfs

username

Username to connect to the storage device

vsadmin

password

Password to connect to the storage device

secret

aggregate

Aggregate for provisioning (optional; if set, must be assigned to the SVM). For the ontap-nas-flexgroup driver, this option is ignored. All aggregates assigned to the SVM are used to provision a FlexGroup Volume.

aggr1

limitAggregateUsage

Optional, fail provisioning if usage is above this percentage

75%

nfsMountOptions

Fine grained control of NFS mount options; defaults to “-o nfsvers=3”. Available only for the ontap-nas and ontap-nas-economy drivers. See NFS host configuration information here.

-o nfsvers=4

igroupName

The igroup used by the plugin; defaults to “netappdvp”. Available only for the `ontap-san`driver.

myigroup

limitVolumeSize

Maximum requestable volume size and qtree parent volume size. For the ontap-nas-economy driver, this option additionally limits the size of the FlexVols that it creates.

300g

qtreesPerFlexvol

Maximum qtrees per FlexVol, must be in range [50, 300], default is 200. For the ontap-nas-economy driver, this option allows customizing the maximum number of qtrees per FlexVol.

300

Default options are available to avoid having to specify them on every volume you create:

Option Description Example

spaceReserve

Space reservation mode; “none” (thin provisioned) or “volume” (thick)

none

snapshotPolicy

Snapshot policy to use, default is “none”

none

snapshotReserve

Snapshot reserve percentage, default is “” to accept ONTAP’s default

10

splitOnClone

Split a clone from its parent upon creation, defaults to “false”

false

encryption

Enables NetApp Volume Encryption (NVE) on the new volume; defaults to false. NVE must be licensed and enabled on the cluster to use this option.

If NAE is enabled on the backend, any volume provisioned in Astra Trident will be NAE enabled.

For more information, refer to: How Astra Trident works with NVE and NAE.

true

unixPermissions

NAS option for provisioned NFS volumes, defaults to “777”

777

snapshotDir

NAS option for access to the .snapshot directory, defaults to “false”

true

exportPolicy

NAS option for the NFS export policy to use, defaults to “default”

default

securityStyle

NAS option for access to the provisioned NFS volume, defaults to “unix”

mixed

fileSystemType

SAN option to select the file system type, defaults to “ext4”

xfs

tieringPolicy

Tiering policy to use, default is “none”; “snapshot-only” for pre-ONTAP 9.5 SVM-DR configuration

none

Scaling options

The ontap-nas and ontap-san drivers create an ONTAP FlexVol for each Docker volume. ONTAP supports up to 1000 FlexVols per cluster node with a cluster maximum of 12,000 FlexVols. If your Docker volume requirements fit within that limitation, the ontap-nas driver is the preferred NAS solution due to the additional features offered by FlexVols, such as Docker-volume-granular snapshots and cloning.

If you need more Docker volumes than can be accommodated by the FlexVol limits, choose the ontap-nas-economy or the ontap-san-economy driver.

The ontap-nas-economy driver creates Docker volumes as ONTAP Qtrees within a pool of automatically managed FlexVols. Qtrees offer far greater scaling, up to 100,000 per cluster node and 2,400,000 per cluster, at the expense of some features. The ontap-nas-economy driver does not support Docker-volume-granular snapshots or cloning.

Note The ontap-nas-economy driver is not currently supported in Docker Swarm, because Swarm does not orchestrate volume creation across multiple nodes.

The ontap-san-economy driver creates Docker volumes as ONTAP LUNs within a shared pool of automatically managed FlexVols. This way, each FlexVol is not restricted to only one LUN and it offers better scalability for SAN workloads. Depending on the storage array, ONTAP supports up to 16384 LUNs per cluster. Because the volumes are LUNs underneath, this driver supports Docker-volume-granular snapshots and cloning.

Choose the ontap-nas-flexgroup driver to increase parallelism to a single volume that can grow into the petabyte range with billions of files. Some ideal use cases for FlexGroups include AI/ML/DL, big data and analytics, software builds, streaming, file repositories, and so on. Trident uses all aggregates assigned to an SVM when provisioning a FlexGroup Volume. FlexGroup support in Trident also has the following considerations:

  • Requires ONTAP version 9.2 or greater.

  • As of this writing, FlexGroups only support NFS v3.

  • Recommended to enable the 64-bit NFSv3 identifiers for the SVM.

  • The minimum recommended FlexGroup size is 100GB.

  • Cloning is not supported for FlexGroup Volumes.

For information about FlexGroups and workloads that are appropriate for FlexGroups see the NetApp FlexGroup Volume Best Practices and Implementation Guide.

To get advanced features and huge scale in the same environment, you can run multiple instances of the Docker Volume Plugin, with one using ontap-nas and another using ontap-nas-economy.

Example ONTAP configuration files

NFS example for ontap-nas driver

{
    "version": 1,
    "storageDriverName": "ontap-nas",
    "managementLIF": "10.0.0.1",
    "dataLIF": "10.0.0.2",
    "svm": "svm_nfs",
    "username": "vsadmin",
    "password": "secret",
    "aggregate": "aggr1",
    "defaults": {
      "size": "10G",
      "spaceReserve": "none",
      "exportPolicy": "default"
    }
}

NFS example for ontap-nas-flexgroup driver

{
    "version": 1,
    "storageDriverName": "ontap-nas-flexgroup",
    "managementLIF": "10.0.0.1",
    "dataLIF": "10.0.0.2",
    "svm": "svm_nfs",
    "username": "vsadmin",
    "password": "secret",
    "defaults": {
      "size": "100G",
      "spaceReserve": "none",
      "exportPolicy": "default"
    }
}

NFS example for ontap-nas-economy driver

{
    "version": 1,
    "storageDriverName": "ontap-nas-economy",
    "managementLIF": "10.0.0.1",
    "dataLIF": "10.0.0.2",
    "svm": "svm_nfs",
    "username": "vsadmin",
    "password": "secret",
    "aggregate": "aggr1"
}

iSCSI example for ontap-san driver

{
    "version": 1,
    "storageDriverName": "ontap-san",
    "managementLIF": "10.0.0.1",
    "dataLIF": "10.0.0.3",
    "svm": "svm_iscsi",
    "username": "vsadmin",
    "password": "secret",
    "aggregate": "aggr1",
    "igroupName": "myigroup"
}

NFS example for ontap-san-economy driver

{
    "version": 1,
    "storageDriverName": "ontap-san-economy",
    "managementLIF": "10.0.0.1",
    "dataLIF": "10.0.0.3",
    "svm": "svm_iscsi_eco",
    "username": "vsadmin",
    "password": "secret",
    "aggregate": "aggr1",
    "igroupName": "myigroup"
}

Element software configuration

In addition to the global configuration values, when using Element software (NetApp HCI/SolidFire), these options are available.

Option Description Example

Endpoint

https://<login>:<password>@<mvip>/json-rpc/<element-version>;

https://admin:admin@192.168.160.3/json-rpc/8.0

SVIP

iSCSI IP address and port

10.0.0.7:3260

TenantName

SolidFireF Tenant to use (created if not found)

“docker”

InitiatorIFace

Specify interface when restricting iSCSI traffic to non-default interface

“default”

Types

QoS specifications

See example below

LegacyNamePrefix

Prefix for upgraded Trident installs. If you used a version of Trident prior to 1.3.2 and perform an upgrade with existing volumes, you’ll need to set this value to access your old volumes that were mapped via the volume-name method.

“netappdvp-”

The solidfire-san driver does not support Docker Swarm.

Example Element software configuration file

{
    "version": 1,
    "storageDriverName": "solidfire-san",
    "Endpoint": "https://admin:admin@192.168.160.3/json-rpc/8.0",
    "SVIP": "10.0.0.7:3260",
    "TenantName": "docker",
    "InitiatorIFace": "default",
    "Types": [
        {
            "Type": "Bronze",
            "Qos": {
                "minIOPS": 1000,
                "maxIOPS": 2000,
                "burstIOPS": 4000
            }
        },
        {
            "Type": "Silver",
            "Qos": {
                "minIOPS": 4000,
                "maxIOPS": 6000,
                "burstIOPS": 8000
            }
        },
        {
            "Type": "Gold",
            "Qos": {
                "minIOPS": 6000,
                "maxIOPS": 8000,
                "burstIOPS": 10000
            }
        }
    ]
}

Cloud Volumes Service (CVS) on GCP configuration

Trident now includes support for smaller volumes with the default CVS service type on GCP. For backends created with storageClass=software, volumes will now have a minimum provisioning size of 300 GiB. NetApp recommends customers consume sub-1TiB volumes for non-production workloads. CVS currently provides this feature under Controlled Availability and does not provide technical support.

Note Sign up for access to sub-1TiB volumes here.
Warning When deploying backends using the default CVS service type storageClass=software, you should obtain access to the sub-1TiB volumes feature on GCP for the Project Number(s) and Project ID(s) in question. This is necessary for Trident to provision sub-1TiB volumes. If not, volume creations will fail for PVCs that are <600 GiB. Obtain access to sub-1TiB volumes using this form.

Volumes created by Trident for the default CVS service level will be provisioned as follows:

  • PVCs that are smaller than 300 GiB will result in Trident creating a 300 GiB CVS volume.

  • PVCs that are between 300 GiB to 600 GiB will result in Trident creating a CVS volume of the requested size.

  • PVCs that are between 600 GiB and 1 TiB will result in Trident creating a 1TiB CVS volume.

  • PVCs that are greater than 1 TiB will result in Trident creating a CVS volume of the requested size.

In addition to the global configuration values, when using CVS on GCP, these options are available.

Option Description Example

apiRegion

CVS account region (required). Is the GCP region where this backend will provision volumes.

“us-west2”

projectNumber

GCP project number (required). Can be found in the GCP web portal’s Home screen.

“123456789012”

hostProjectNumber

GCP shared VPC host project number (required if using a shared VPC)

“098765432109”

apiKey

API key for GCP service account with CVS admin role (required). Is the JSON-formatted contents of a GCP service account’s private key file (copied verbatim into the backend config file). The service account must have the netappcloudvolumes.admin role.

(contents of the private key file)

secretKey

CVS account secret key (required). Can be found in the CVS web portal in Account settings > API access.

“default”

proxyURL

Proxy URL if proxy server required to connect to the CVS account. The proxy server can either be an HTTP proxy or an HTTPS proxy. In case of an HTTPS proxy, certificate validation is skipped to allow the usage of self-signed certificates in the proxy server. Proxy servers with authentication enabled are not supported.

“http://proxy-server-hostname/”

nfsMountOptions

NFS mount options; defaults to “-o nfsvers=3”

“nfsvers=3,proto=tcp,timeo=600”

serviceLevel

Performance level (standard, premium, extreme), defaults to “standard”

“premium”

network

GCP network used for CVS volumes, defaults to “default”

“default”
Note If using a shared VPC network, you should specify both projectNumber and hostProjectNumber. In that case, projectNumber is the service project and hostProjectNumber is the host project.
Note The NetApp Cloud Volumes Service for GCP does not support CVS-Performance volumes less than 100 GiB in size, or CVS volumes less than 300 GiB in size. To make it easier to deploy applications, Trident automatically creates volumes of the minimum size if a too-small volume is requested.

When using CVS on GCP, these default volume option settings are available.

Option Description Example

exportRule

NFS access list (addresses and/or CIDR subnets), defaults to “0.0.0.0/0”

“10.0.1.0/24,10.0.2.100”

snapshotDir

Controls visibility of the .snapshot directory

“false”

snapshotReserve

Snapshot reserve percentage, default is “” to accept the CVS default of 0

“10”

size

Volume size, defaults to “100GiB”

“10T”

Example CVS on GCP configuration file

{
    "version": 1,
    "storageDriverName": "gcp-cvs",
    "projectNumber": "012345678901",
    "apiRegion": "us-west2",
    "apiKey": {
        "type": "service_account",
        "project_id": "my-gcp-project",
        "private_key_id": "<id_value>",
        "private_key": "
        -----BEGIN PRIVATE KEY-----
        <key_value>
        -----END PRIVATE KEY-----\n",
        "client_email": "cloudvolumes-admin-sa@my-gcp-project.iam.gserviceaccount.com",
        "client_id": "123456789012345678901",
        "auth_uri": "https://accounts.google.com/o/oauth2/auth",
        "token_uri": "https://oauth2.googleapis.com/token",
        "auth_provider_x509_cert_url": "https://www.googleapis.com/oauth2/v1/certs",
        "client_x509_cert_url": "https://www.googleapis.com/robot/v1/metadata/x509/cloudvolumes-admin-sa%40my-gcp-project.iam.gserviceaccount.com"
    },
    "proxyURL": "http://proxy-server-hostname/"
}

Azure NetApp Files configuration

To configure and use an Azure NetApp Files backend, you will need the following:

  • subscriptionID from an Azure subscription with Azure NetApp Files enabled

  • tenantID, clientID, and clientSecret from an App Registration in Azure Active Directory with sufficient permissions to the Azure NetApp Files service

  • Azure location that contains at least one delegated subnet

Tip If you’re using Azure NetApp Files for the first time or in a new location, some initial configuration is required that the quickstart guide will walk you through.
Note Astra Trident 21.04.0 and earlier do not support Manual QoS capacity pools.
Option Description Default

version

Always 1

storageDriverName

“azure-netapp-files”

backendName

Custom name for the storage backend

Driver name + “_” + random characters

subscriptionID

The subscription ID from your Azure subscription

tenantID

The tenant ID from an App Registration

clientID

The client ID from an App Registration

clientSecret

The client secret from an App Registration

serviceLevel

One of “Standard”, “Premium” or “Ultra”

“” (random)

location

Name of the Azure location new volumes will be created in

“” (random)

virtualNetwork

Name of a virtual network with a delegated subnet

“” (random)

subnet

Name of a subnet delegated to Microsoft.Netapp/volumes

“” (random)

nfsMountOptions

Fine-grained control of NFS mount options

“-o nfsvers=3”

limitVolumeSize

Fail provisioning if requested volume size is above this value

“” (not enforced by default)
Note The Azure NetApp Files service does not support volumes less than 100 GB in size. To make it easier to deploy applications, Trident automatically creates 100 GB volumes if a smaller volume is requested.

You can control how each volume is provisioned by default using these options in a special section of the configuration.

Option Description Default

exportRule

The export rule(s) for new volumes. Must be a comma-separated list of any combination of IPv4 addresses or IPv4 subnets in CIDR notation.

“0.0.0.0/0”

snapshotDir

Controls visibility of the .snapshot directory

“false”

size

The default size of new volumes

“100G”

Example Azure NetApp Files configurations

Example 1: Minimal backend configuration for azure-netapp-files

This is the absolute minimum backend configuration. With this configuration, Trident will discover all of your NetApp accounts, capacity pools, and subnets delegated to ANF in every location worldwide, and place new volumes on one of them randomly.

This configuration is useful when you’re just getting started with ANF and trying things out, but in practice you’re going to want to provide additional scoping for the volumes you provision to make sure that they have the characteristics you want and end up on a network that’s close to the compute that’s using it. See the subsequent examples for more details.

{
    "version": 1,
    "storageDriverName": "azure-netapp-files",
    "subscriptionID": "9f87c765-4774-fake-ae98-a721add45451",
    "tenantID": "68e4f836-edc1-fake-bff9-b2d865ee56cf",
    "clientID": "dd043f63-bf8e-fake-8076-8de91e5713aa",
    "clientSecret": "SECRET"
}

Example 2: Single location and specific service level for azure-netapp-files

This backend configuration places volumes in Azure’s “eastus” location in a “Premium” capacity pool. Trident automatically discovers all of the subnets delegated to ANF in that location and will place a new volume on one of them randomly.

{
    "version": 1,
    "storageDriverName": "azure-netapp-files",
    "subscriptionID": "9f87c765-4774-fake-ae98-a721add45451",
    "tenantID": "68e4f836-edc1-fake-bff9-b2d865ee56cf",
    "clientID": "dd043f63-bf8e-fake-8076-8de91e5713aa",
    "clientSecret": "SECRET",
    "location": "eastus",
    "serviceLevel": "Premium"
}

Example 3: Advanced configuration for azure-netapp-files

This backend configuration further reduces the scope of volume placement to a single subnet, and also modifies some volume provisioning defaults.

{
    "version": 1,
    "storageDriverName": "azure-netapp-files",
    "subscriptionID": "9f87c765-4774-fake-ae98-a721add45451",
    "tenantID": "68e4f836-edc1-fake-bff9-b2d865ee56cf",
    "clientID": "dd043f63-bf8e-fake-8076-8de91e5713aa",
    "clientSecret": "SECRET",
    "location": "eastus",
    "serviceLevel": "Premium",
    "virtualNetwork": "my-virtual-network",
    "subnet": "my-subnet",
    "nfsMountOptions": "nfsvers=3,proto=tcp,timeo=600",
    "limitVolumeSize": "500Gi",
    "defaults": {
        "exportRule": "10.0.0.0/24,10.0.1.0/24,10.0.2.100",
        "size": "200Gi"
    }
}

Example 4: Virtual storage pools with azure-netapp-files

This backend configuration defines multiple pools of storage in a single file. This is useful when you have multiple capacity pools supporting different service levels and you want to create storage classes in Kubernetes that represent those.

This is just scratching the surface of the power of virtual storage pools and their labels.

{
    "version": 1,
    "storageDriverName": "azure-netapp-files",
    "subscriptionID": "9f87c765-4774-fake-ae98-a721add45451",
    "tenantID": "68e4f836-edc1-fake-bff9-b2d865ee56cf",
    "clientID": "dd043f63-bf8e-fake-8076-8de91e5713aa",
    "clientSecret": "SECRET",
    "nfsMountOptions": "nfsvers=3,proto=tcp,timeo=600",
    "labels": {
        "cloud": "azure"
    },
    "location": "eastus",

    "storage": [
        {
            "labels": {
                "performance": "gold"
            },
            "serviceLevel": "Ultra"
        },
        {
            "labels": {
                "performance": "silver"
            },
            "serviceLevel": "Premium"
        },
        {
            "labels": {
                "performance": "bronze"
            },
            "serviceLevel": "Standard",
        }
    ]
}