Planning for high availability
You must plan which system topology to use and if you intend to link with partner systems at other sites.
Before you begin
Choose one of the following system topologies:
- The standard topology, where all nodes are at
the same site. Higher availability can be achieved by using Metro Mirror or Global Mirror to maintain a
copy of a volume on a different system at a remote site. Each node in an I/O
group can be placed at a different site. This topology is also known as a standard stretched system.
For information about planning for a standard stretched system, see Stretched system
configuration details in the Related concepts section.
For information about planning for Metro Mirror and Global Mirror, see Configuring IP partnerships in the Related concepts section.
- The stretched topology, where each node of an I/O group is at a
different site and volumes have a copy at each site. Access to a volume can continue when one site
is not available but with reduced performance. This topology is also known as an enhanced stretched
system.
The reason for using an enhanced stretched system rather than the HyperSwap® topology is either because of using Global Mirror or Metro Mirror to a third site, or because the system was configured as an enhanced stretched system before HyperSwap being released.
For information about planning for an enhanced stretched system, see Planning for a stretched topology system
- The HyperSwap topology where each I/O group and its managed disks are at a different site. Active-active remote copy relationships mean that a volume can be active on two I/O groups so that it can immediately be accessed by using the other site without loss of performance. For information about planning for a HyperSwap topology system, see Planning for a HyperSwap topology system.