Storage Virtual Machines (SVMs, formerly known as Vservers) provide data access to clients without regard to physical storage or controller, similar to any storage system. When you use SVMs, they provide benefits such as nondisruptive operation, scalability, security and support unified storage.
SVMs have the following benefits:
• Nondisruptive operation
SVMs can operate continuously and nondisruptively for as long as they are needed. SVMs help clusters to operate continuously during software and hardware upgrades, addition and removal of nodes, and all administrative operations.
• Scalability
SVMs meet on-demand data throughput and the other storage requirements.
• Security
Each SVM appears as a single independent server, which enables multiple SVMs to coexist while ensuring no data flows among them.
• Unified storage
SVMs can serve data concurrently through multiple data access protocols. SVMs provide file-level data access by using NAS protocols, such as CIFS and NFS, and block-file-level data access by using SAN protocols, such as iSCSI and FC (FCoE included). SVMs can serve data to SAN and NAS clients independently at the same time.
Note: SVMs with Infinite Volume can serve data only through NFS and CIFS (SMB 1.0) protocols.
• Delegation of management
Each SVM can have its own user and administration authentication. SVM administrators can manage the SVMs that they are authorized to access. However, SVM administrators have privileges assigned by the cluster administrators.
• Easy management of large datasets
With SVMs with Infinite Volume, management of large and unstructured data is easier as the SVM administrator has to manage one data container instead of many.
Managing SVMs
Creating SVMs with FlexVol volumes on Data ONTAP 8.1.x
You can create Storage Virtual Machines (SVMs) with FlexVol volumes to serve data to the clients.
Before you begin
• The cluster must have at least one non-root aggregate.
• You must have synchronized the time across the cluster by configuring and enabling NTP to prevent CIFS creation and authentication failures.
• SVMs with Infinite Volume must not exist in the cluster.
• Protocols that you want to configure or allow on the SVM must be licensed.
About this task
You cannot use System Manager to create SVMs with Infinite Volume on storage systems running Data ONTAP 8.1.x.
You can create SVMs and perform the following tasks by using the Create Storage Virtual Machine wizard:
• Delegate SVM management to SVM administrators.
• Set up network interfaces for the SVM.
You can allow management access to a data LIF if you delegate SVM management to SVM administrators.
• Configure services.
• Configure protocols.
Data ONTAP creates certain UNIX groups and users, which are local to the SVM. The default local groups are daemon with GID (Group ID) 1, pcuser with GID 65534, and nobody with GID 65534.
The default local users are root with UID (User ID) 0, pcuser with UID 65534, and nobody with UID 65534.
Note: You cannot use these default names while creating local users and groups.
Steps
1. From the home page, double-click the appropriate storage system.
2. Expand the Storage Virtual Machines hierarchy in the left navigation pane.
3. In the navigation pane, select the cluster.
4. Click Create.
5. Select the SVM type as Storage Virtual Machine with FlexVol volume and click Next.
This option is available only if the cluster does not contain any SVM.
6. Type or select information as prompted by the wizard.
7. Confirm the details and click Finish to complete the wizard.
8. Click Close after the SVM setup is complete.
Result
The SVM is created with a root volume with UNIX style security and UNIX 700 “read write execute” permissions for the Owner. The root volume name is automatically generated as SVM name_root.
The SVM that you created is started automatically. If you delegated the administration to an SVM administrator, the vsadmin user account is created and is in an unlocked state. The vsadmin role is assigned to the default vsadmin user account.
Creating SVMs on Data ONTAP 8.2.x
You can create fully configured Storage Virtual Machines (SVMs) that can serve data immediately.
A cluster can have one or more SVMs with FlexVol volumes and SVMs with Infinite Volume.
Before you begin
• The cluster must have at least one non-root aggregate in the online state.
• The aggregate must have sufficient space for the SVM root volume.
• You must have synchronized the time across the cluster by configuring and enabling NTP to prevent CIFS creation and authentication failures.
• Protocols that you want to configure on the SVM must be licensed.
About this task
• While creating SVMs, you can perform the following tasks:
◦ Create and fully configure SVMs.
◦ Configure the volume type allowed on SVMs.
◦ Create and configure SVMs with minimal network configuration.
◦ Delegate the administration to SVM administrators.
• To name the SVM, you can use alphanumeric characters and the following special characters: "."
(period), "-" (hyphen), and "_" (underscore).
The SVM name should start with an alphabet or "_" (underscore) and cannot exceed 47 characters in length.
Note: You should use unique fully qualified domain names (FQDN) for the SVM name such as vs0.example.com.
• You can establish SnapMirror relationships only between volumes that have the same language settings.
The language of the SVM determines the character set that is used to display file names and data for all NAS volumes in the SVM.
Steps
1. From the home page, double-click the appropriate storage system.
2. Expand the Storage Virtual Machines hierarchy in the left navigation pane.
3. In the navigation pane, select the cluster.
4. Click Create.
5. In the Storage Virtual Machine (SVM) Setup window, specify the SVM details such as the SVM name, volume type allowed, protocols allowed, SVM language, root volume security style, and its root aggregate.
The default language setting for an SVM is C.UTF-8.
By default, the aggregate with the maximum free space is selected as the container for root volume of the SVM. Based on the protocols selected, the default security style and the root aggregate are selected. The security style is set to NTFS if you select CIFS protocol or a combination of CIFS protocol with the other protocols. The security style is set to Unix if you select NFS, iSCSI, or FC/FCoE or a combination of these protocols.
6. Specify the DNS domain names and the name server IP addresses to configure the DNS services.
The default values are selected from the existing SVM configurations.
7. Click Submit & Continue.
The SVM is created with the specified configuration.
Result
The SVM that you created is started automatically. The root volume name is automatically generated as SVM name_root. By default, the vsadmin user account is created and is in the locked state.
After you finish
• You must configure at least one protocol on the SVM to allow data access.
• After you create an SVM with Infinite Volume, you must create an Infinite Volume for the SVM.
Configuring CIFS and NFS protocols on an SVM
You can configure CIFS and NFS protocols on the Storage Virtual Machine (SVM) to provide file-level data access for NAS clients. To enable CIFS protocol, you must create the data LIFs and the CIFS server. To enable NFS protocol, you can specify the NIS details in addition to the data LIFs.
Before you begin
• Protocols that you want to configure or allow on the SVM must be licensed.
If the protocol is not allowed on the SVM, you can use the Edit Storage Virtual Machine window to enable the protocol for the SVM.
• You must have the Active Directory, Organizational unit, and administrative account credentials for configuring CIFS protocol
• You must have the IP address, netmask, and gateway information to create the IP network interface on the SVM.
Steps
1. If you have not configured the protocols while creating the SVM, double-click the appropriate storage system in the home page, expand the Storage Virtual Machines hierarchy, select the SVM in the navigation pane, and click the protocol you want to configure.
2. In the Data LIF Configuration section, specify the network details to create data LIFs.
You can either retain the same data LIF configuration for both CIFS and NFS or configure a new LIF for each protocol.
3. Specify the following information to create a CIFS server.
• CIFS server name
• Active Directory to associate with the CIFS server
• Organizational unit (OU) within the Active Directory domain to associate with the CIFS server
By default, this parameter is set to CN=Computers.
• Credentials of an administrative account that has sufficient privileges to add the CIFS server to the OU
4. Optional: You can also specify the IP addresses of the NIS servers and NIS domain name to configure NIS services on the SVM.
5. Click Submit & Close.
Result
The CIFS server and NIS domain are configured with the specified configuration. Data LIFs are created. By default, the data LIFs have management access. You can view the configuration details on the Summary page.
Configuring iSCSI protocol on SVMs
You can configure the iSCSI protocol on the Storage Virtual Machine (SVM) to provide block-level data access. You can create iSCSI LIFs and portsets, and add the LIFs to the portsets. LIFs are created on the most suitable adapters and assigned to portsets to ensure data path redundancy.
Before you begin
• The iSCSI license must be enabled on the cluster.
If the protocol is not allowed on the SVM, you can use the Edit Storage Virtual Machine window to enable the protocol for the SVM.
• All the nodes in the cluster must be healthy.
• Each node must have at least two data ports and the port state must be up. About this task
You can configure the iSCSI protocol while creating the SVM or you can do so at a later time.
Steps
1. If you have not configured the iSCSI protocol while creating the SVM and if the FCP service is not started, double-click the appropriate storage system in the home page, expand the Storage Virtual Machines hierarchy, select the SVM in the navigation pane, and click iSCSI.
2. Optional: In the Data Interface Configuration section, specify an alias for the iSCSI target.
The maximum number of characters for an alias name is 128. If you do not specify a target alias, the SVM name is used as an alias.
3. Specify the number of iSCSI LIFs that can be assigned to a single node.
For storage systems running Data ONTAP 8.2 or earlier, the minimum number of LIFs per node is two. Starting from Data ONTAP 8.2.1, the minimum number of LIFs per node is one. The maximum number is the minimum of all the ports in the up state across the nodes. If the maximum value is an odd number, the previous even number is considered as the maximum value. You can choose any even number in the minimum and maximum value range.
Note: Do not manually enter the value. You must the select a value from the list.
Example
A 4-node cluster has node1, node2, and node3 with 6 ports each in the up state, and node4 with 7 ports in the up state. The effective maximum value for the cluster is 6.
4. Specify the network details to create iSCSI LIFs.
The starting IP address specifies the first of the contiguous addresses in the LIF IP address pool.
5. If you want to verify or modify the automatically generated iSCSI LIFs configuration, select Review or Modify LIFs configuration (Advanced Settings).
You can modify only the LIF name, home port, and LIF IP address. By default, the portsets are set to the minimum value. You must ensure that you do not specify duplicate entries. If you specify duplicate LIF names, System Manager appends numeric values to the duplicate LIF name.
Based on the selected portset, the LIFs are distributed across the portsets using a round-robin method to ensure redundancy in case of node or port failure.
6. Click Submit & Continue.
Result
The iSCSI LIFs and portsets are created with the specified configuration. The LIFs are distributed accordingly among the portsets. The iSCSI service is started if all the LIFs are successfully created.
If the LIF creation fails, you can use the Network Interfaces window to create the LIFs, attach the LIFs to the portsets by using the LUNs window, and start the iSCSI service by using the iSCSI window.
Configuring FC and FCoE protocols on SVMs
You can configure the FC and the FCoE protocols on the SVM for SAN hosts. LIFs are created on the most suitable adapters and assigned to port sets to ensure data path redundancy. Based on your requirements, you can configure either FC, FCoE, or both the protocols.
Before you begin
• The FCP license must be enabled on the cluster.
• All the nodes in the cluster must be healthy.
• Each node must have at least two correctly configured ports for each protocol (FC and FCoE).
About this task
You can configure the FC and FCoE protocols while creating the SVM or you can do so at a later time. If the protocols are not allowed on the SVM, you can use the Edit Storage Virtual Machine window to enable the protocols for the SVM.
For information about managing and configuring SAN FC and FCoE environments, see the Clustered Data ONTAP SAN Administration Guide.
Steps
1. If you have not configured the protocols while creating the SVM and if the FCP is not started, double-click the appropriate storage system in the home page, expand the Storage Virtual Machines hierarchy, select the SVM in the navigation pane, and click FC/FCoE.
2. In the Data Interface Configuration section, select the corresponding option to configure data LIFs for FC and FCoE protocols.
3. Specify the number of data LIFs per node for each protocol.
For storage systems running Data ONTAP 8.2 or earlier, the minimum number of LIFs per node is two. Starting from Data ONTAP 8.2.1, the minimum number of LIFs per node is one. The maximum number is the minimum of all the ports in the up state across the nodes. If the maximum value is an odd number, the previous even number is considered as the maximum value. You can choose any even number in the minimum and maximum value range.
Note: Do not manually enter the value. You must the select a value from the list.
Example
A 4-node cluster has node1, node2, and node3 with 6 ports each in the up state, and node4 with 7 ports in the up state. The effective maximum value for the cluster is 6.
4. If you want to verify or modify the automatically generated LIFs configuration, select Review or Edit the Interface Association.
You can modify only the LIF name and home port. You must ensure that you do not specify duplicate entries.
5. Click Submit & Continue.
Result
The data LIFs and port sets are created with the specified configuration. The LIFs are distributed accordingly among the port sets. FCP service is started if all the LIFs are successfully created for at least one protocol.
If the LIF creation fails, you can create the LIFs and start the FCP service from the FC/FCoE window.
Delegating administration to SVM administrators
After setting up a functional Storage Virtual Machine (SVM) or SVMs with basic network configuration, you can optionally delegate the administration of the SVM to SVM administrators.
Steps
1. In the Administrator Details section, set up a password for the vsadmin user account.
2. If you want a dedicated LIF for SVM management, select Create a LIF for SVM management and specify the network details.
A dedicated SVM management LIF is required for SAN protocols, where data and management protocols cannot share the same LIF. SVM management LIFs can be created only on data ports.
Result
The vsadmin account is unlocked and configured with the password.
The default access methods for the vsadmin account are Data ONTAP API (ontapi) and SSH (ssh). The SVM administrator can log in to the storage system using the management IP address.
After you finish
You must assign aggregates to the SVM by using the Edit Storage Virtual Machine dialog box.
Note: If the SVM does not have any assigned aggregates, the SVM administrator cannot create volumes.
Assigning aggregates to SVMs
After creating an SVM for an Infinite Volume, you should assign specific aggregates to it so that the Infinite Volume that you create will use those specific aggregates and not use all the aggregates in the cluster.
Before you begin
You should have reviewed the available aggregates and decided which aggregates the SVM will use.
About this task
You identify which aggregates the Infinite Volume will use by assigning aggregates to its containing SVM with Infinite Volume. If you do not specify the aggregate list for the SVM with Infinite Volume, the Infinite Volume can potentially use all the aggregates in the cluster.
Steps
1. In the Select aggregates section, select the aggregates to assign to the SVM.
By default, the node root aggregates are not selected. You should not provision volumes on root aggregates because it might cause performance or stability issues.
2. Click Submit & Continue.
Editing SVM settings
You can edit the properties of Storage Virtual Machines (SVMs), such as the name service switch, name mapping switch, and aggregate list.
About this task
You can edit the values of the following properties of SVMs:
• Name service switch
• Name mapping switch
• Protocols that are allowed to serve data
Note: The CIFS protocol that is configured on the SVM continues to serve data even when you disallow it on that SVM.
• The list of the aggregates that are available to create volumes
If you do not specify the aggregates for SVMs with Infinite Volume, the Infinite Volume spans across all the aggregates in the cluster.
Note: For FlexVol volumes, you can assign aggregates only if you have delegated administration to an SVM administrator.
Steps
1. From the home page, double-click the appropriate storage system.
2. Expand the Storage Virtual Machines hierarchy in the left navigation pane.
3. In the navigation pane, select the appropriate cluster.
4. Select the SVM that you want to edit and click Edit.
5. Make the necessary changes and click Save and Close.
Deleting SVMs
You can delete Storage Virtual Machines (SVMs) that you no longer require from the storage system configuration.
Before you begin
You must have completed the following tasks:
1. Disabled the Snapshot copies, data protection (DP) mirrors, and load-sharing (LS) mirrors for all the volumes
Note: You must use the CLI to disable LS mirrors.
2. Deleted all the igroups that belong to the SVM manually if you are deleting SVMs with FlexVol volume
3. Deleted all the portsets
4. Deleted all the volumes in the SVM, including the root volume 5. Unmapped the LUNs, taken them offline, and deleted them
4. Deleted all the volumes in the SVM, including the root volume 5. Unmapped the LUNs, taken them offline, and deleted them