Specifications

© IBM Copyright, 2012 Version: January 26, 2012
www.ibm.com/support/techdocs 24
Summary of Best Practices for Storage Area Networks
unable to communicate to your disk arrays or mirror write cache because you have a
single congested link leading to an edge switch.
If at all possible, plan and design for the maximum size configuration the SVC
cluster could eventually reach. The design of the SAN can change radically for larger
numbers of hosts. Modifying the SAN later to accommodate a larger-than-expected
number of hosts will either produce a poorly-designed SAN or be very difficult,
expensive, and disruptive. This does not mean you need to purchase all of the SAN
hardware initially, just that you need to layout the SAN while keeping the maximum
size in mind. Always deploy at least one “extra” ISL per switch. Not doing so may
lead to consequences from complete path loss (this is bad), to fabric congestion (this
is even worse).
With the exception of a Stretched Cluster SVC solution, never split the two nodes in
an I/O group between different switches. However, in a dual-fabric configuration, half
the nodes ports must remain on the same switch with the other half of the ports on
another switch.
Unless two clusters participate in a mirroring relationship, all zoning should be
configured so that the two clusters do not share a zone. If a single host requires
access to two different clusters, create two zones, each to a separate cluster. The
storage zones should also be separate, even if the two clusters share a disk
controller.
If a FlashCopy mapping is created, where the source VDisk is a target VDisk of an
active Metro Mirror relationship, then this will add additional latency to that existing
Metro Mirror relationship (and possibly affect the hosts which are using the source
VDisk of that Metro Mirror relationship as a result). The reason for the additional
latency, is that the FlashCopy prepare disables the cache on the source VDisk
(which is the target VDisk of the Metro Mirror relationship), and thus all write I/Os
from the Metro Mirror relationship need to commit to the storage controller, before
the complete is returned to the host.
When connecting an SVC to a DS8000, there are several standard configurations
that should be used:
4-adapter/8-port configuration: 32 array max
4-adapter/16-port configuration: 48 array max (marginal configuration)
8-adapter/16-port configuration: >48 array max (or multiple SVC clusters)