Understanding and Designing Serviceguard Disaster Recovery Architectures

You can use SONET instead of DWDM in this configuration by replacing the DWDM boxes with
SONET boxes. With SONET, you can use a point to point topology, as is shown in this example,
or you can use a dual SONET ring topology which passes through all three data centers, where
the working ring and the protection ring are alternately routed.
Figure 28 Extended Cluster for RAC with Two Data Centers and a Third Location with WDM Links
Quorum server - Third Location
Primary Data Center 2Primary Data Center 1
QS Node
FC Switch
This network is
used for RAC
cache
fusion traffic.
Node A
Node B
Node C Node D
Maximum distance
100 kilometers
DWDM
FC Switch
FC Switch
Network Switch 1 Network Switch 5
Network
Switch 4
Network Switch 2
Network
Switch 3
FC Switch
FC Switch
Storage Storage
Redundant DWDM
DWDM
WAN
Network Switch 6
This example, Figure 28 (page 66) is specifically for an EC RAC cluster. The WDM boxes connected
between the two Primary Data Centers are configured with redundant fiber links and the standby
fiber feature is enabled. Note that there is a separate network (indicated by the green lines to
switches 3 and 4) being used for the RAC Cache Fusion traffic to ensure good RAC performance.
Switches 2 and 5 are used for the Standby network, which can provide local LAN failover for both
the Primary Heartbeat network and the Primary RAC Cache Fusion network. However, the Standby
network can only provide local failover capability for one of the Primary networks at a time.
Therefore, it is preferable to have a separate Standby network for the Heartbeat network and for
the RAC Cache Fusion network. The distance between the Quorum Server Data center and the
Primary Data Centers is not specified, however you must ensure that the Quorum Server can be
contacted within a reasonable amount of time (should be within the NODE_TIMEOUT period for
Serviceguard A.11.18 or before, or MEMBER_TIMEOUT period for Serviceguard A.11.19 and
later). Cluster lock disks are not allowed in this configuration. There can be 2, 4, 6, or 8 nodes
66 Extended Distance Cluster Configurations