Stretched volumes

Stretched volumes create copies on separate sites for systems that are configured with stretched system topology. Stretched volumes create copies that are on sites that were specified when the stretched system topology was configured. Each site maintains an active version of data that is written to simultaneously. If the other site becomes unavailable, then either site can provide access to the data.

Stretched system topology enables disaster recovery and high availability between nodes in I/O groups at different locations. Each node in an I/O group is defined on separate sites. Before you configure stretched volumes, the system topology needs to be configured for stretched topology and sites must be defined.

In the management GUI, stretched volumes are configured by specifying volume details such as quantity, capacity, name, and the method for saving capacity. As with basic volumes, you can choose either compression or thin-provisioning to save capacity on volumes. For thin-provisioning or compression, you can also select to use deduplication for the volume that you create. For example, you can create a compressed volume that also uses deduplication to remove duplicated data. The method for capacity savings applies to all stretched volumes and copies that are created. The volume location displays the site where copies are located, based on the configured sites for the stretched system topology. For each site, specify a pool for the volume copies that are created. If you select to deduplicate volume data, the volume copies must be in data reduction pools on both sites. Select the I/O group that applies to both sites. Each node in the I/O group is assigned to a site when the topology for the stretched system is configured.

In the management GUI, an asterisk (*) indicates the local copy of the stretched volume. The local copy is the volume that is used for read requests. Unlike mirrored volumes, stretched volumes do not have a primary copy that you can change. Each site automatically considers the local copy to be the primary copy to optimize I/O routing.