Sample Configurations with SGeRAC and Oracle RAC 10gR2, March 2009

Figure 3. Single SG-HB with CSS-HB and RAC-DB-IC on separate subnet

Figure 3 is a variation of figure 2 and shows where the CSS-HB residing on the same subnet as SG-

HB. The RAC-DB-IC is on a separate network and thus do not affect the HB traffic. If the primary

(lan1) fails, Serviceguard performs local LAN failover. If both primary (lan1) and standby (lan2) fails,

RAC Instance Membership Recovery (IMR) reforms and evicts suspect nodes. Eviction reboots the

node. The IMR timeout is by default longer than the Serviceguard reconfiguration time and CSS

reconfiguration time.

The advantage of this configuration is that RAC instance traffic is separate from heartbeat traffic. The

RAC traffic does not interfere with heartbeat traffic. A Serviceguard package can be configured to

monitor the RAC-DB-IC subnet. If the RAC-DB-IC subnet fails (both primary and standby), the

Serviceguard package can be configured to shutdown the RAC instance in order to avoid RDBMS

IMR timeout, and only the specific RAC instance is affected. This configuration allows halting the RAC

instance rather than evicting and halting the node. Use of Serviceguard subnet monitoring has a

limitation where if all interconnect fails (for example primary and standby switch failed at the same

time), all the instances are halted. If there is a concern with simultaneous failure of both switches,

starting with SGeRAC A.11.18, Serviceguard supports cluster interconnect subnet monitoring.

Cluster Interconnect Configurations for CFS and CVM

Configurations with CFS and CVM add GAB/LLT as an additional set of traffic.

Node A

LAN 1

Node B

RAC

Private

(primary)

Private

(standby)

Private

(primary)

LAN 1

LAN 2

CSS-HB

Private

(standby)

LAN 2

LAN 4

LAN 3