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

6
§ With RAC traffic, if the interconnect is not configured to be monitored and
acted upon by the other components, RAC discovers the interconnect failure
within the Instance Membership Recovery (IMR) timeout.
The failover time requirement determines important timeouts, such as Serviceguard heartbeat timeout,
network polling intervals, and cluster interconnect monitoring.
Planning for high availability
A properly configured high availability configuration should survive single failures and continue to
operate.
Public network
The following describes the two levels of client public network high availability (HA): redundant
components and client failover.
Redundant network interfaces and switches with local LAN failover by Serviceguard or bonding by
Auto-Port Aggregation (APA) protects against single point network failures.
Client failover is needed with failures that impact existing and new client sessions. These failures
include node failures (such as those caused by a power failure) and network failures (for example all
redundant network interface/link failed). Protection is available at three levels: Oracle Fast
Application Notification (FAN), remote VIP failover, and client connect timeout. Clients that are FAN
integrated or using the FAN API may interrupt existing sessions and failover. Remote VIP failover is
useful for non-FAN clients attempting to connect to the local node to avoid the TCP connect timeout.
The client connect timeout is useful when client connect takes a long time for whatever reason.
VIP high availability
This section describes high availability for VIP.
Note:
Previously, Oracle virtual IP address (VIP) and Serviceguard relocatable IP
address (RIP) should not exist on the same subnet on the same node due to
potential collisions on IP address configuration. This issue has been
addressed in Oracle 10.2.0.2 for Integrity platform and in Oracle
10.2.0.3 for HP 9000 platforms.
Serviceguard local LAN failover mechanism – preferred choice
For client public network HA in a SGeRAC configuration, the preferred method for network HA is to
use Serviceguard primary and standby. Serviceguard monitors the redundant network and additional
APA software is not required.
When the client network is configured with Serviceguard local LAN failover, Serviceguard performs
the local LAN failover and Oracle Clusterware (OC) configures the VIP after Serviceguard local LAN
failover. Since OC performs monitoring and manages the VIP address, client connectivity maybe
unavailable until OC detects the outage and configures the VIP address on the local node
2
.
2
See Doc ID 296874.1 Configuring the HP-UX Operating System for the Oracle 10g VIP at https://metalink.oracle.com/ (Oracle MetaLink
account required)