Failover within a subnet of a GCP zone using virtual private IP - Linux

Last published : Jun 10, 2026
InfoScale clusters let you fail over IPs - and thereby, the application configured for HA - within the same subnet of a zone.
The following information is required:
  • The virtual private IP (secondary private IP) address to be failed over
  • The device to which the IP should be plumbed
  • The google-api-python-client GCP Python module. For details, see the Bundled Agents Reference Guide - Linux.
Sample configuration with private IP
The following graphic depicts a sample failover configuration within the same subnet using a virtual private IP:
image
The sample configuration includes the following elements:
  • A VPC network is configured in Region A of the Google cloud.
  • An application is configured for HA using an InfoScale cluster that comprises two nodes, Node1 and Node2, which are VM instances.
  • Both the cluster nodes exist in the same subnet.
  • A virtual private IP is configured, which is failed over from one node to the other as part of the failover or the failback operations.
Sample service group configuration with private IP
The following snippet is a service group configuration from a sample VCS configuration file (main.cf):
group oragrp (
    SystemList = { cloud-vm1 = 0, cloud-vm2 = 1 }
        )

    GoogleIP gipres (
        Device = eth0
        PrivateIP = "10.209.1.20"
        )

    IP physipres (
        Device = eth0
        Address = "10.209.1.20"
        NetMask = "255.255.255.255"
        )

    NIC nicres (
        Device = eth0
        )

gipres requires physipres
physipres requires nicres