NetApp DataONTAP Cluster Mode Learning Notes 3

DATA SVM

Data SVM is the basic unit of secure multi-tenancy, which can partition the cluster into multiple unrelated SVM. Each SVM-to-client process is now a single independent server.

Unless turned on by the cluster administrator, traffic in the cluster will not be transferred between SVM.

If multi-tenancy is not enabled, at least one DataSVM is still required.

SVM used to be called vserver and SVM in the GUI interface, but the CLI mode still uses the "vserver" command.

There can be only one or more SVM in the cluster, a single SVM can provide SAN or NAS protocols, or both.

Each SVM is a proprietary server for the client

Data SVM is independent of nodes and aggregate, and it can be based on multiple nodes and aggregate, or on only one. Different DataSVM can be on the same node and use the same aggregate. You can also specify SVM to specify an aggregate. You can specify dedicated volumes and Data LIFs for SMV.

Data Namespace

Each SVM has its own unique "namespace" for NAS and its own directory structure.

When a SVM is created, its root volume is also created.

To build namespaces, the volumes in each SVM are connected to each other through junction path and mounted on the junction path.

The root volume is at the top of the namespace hierarchy, and the other volumes are mounted under the root volume through junction path.

IPSpaces

IPspaces allows SVM in the same cluster to have overlapping subnets and IP addresses, and on previous Data ONTAP, Data LIFs had to use different IP addresses.

With IPspaces, LIFs on different SVM can use the same IP address of other SVM.

IPspaces is not necessary for secure multi-tenancy because the address space is already provided by SVM.

Each SVM maintains its own routing table, and IPspaces is equivalent to VRF (Virtual routing forwarding Table).