When Cluster fail-over occurs, even if the HUB monitoring function of GLS detects failures in the entire transfer route on the node to be switched, the Gls resource does not become faulted and stopping userApplication can be suppressed.
With this capability, when entire transfer routes are recovered, communication can be recovered without recovery operation.
This function can be used in a configuration that satisfies all of the following conditions.
1:1 Standby Operation
One userApplication
NIC switching mode
With the exception of the Gls resource, userApplication has no resources to detect network faults.
The application that can restart operation when communication is recovered is used.
See
For information on the setup, see Section "3.12.3 Setting suppression of stopping userApplication when entire transfer routes fails".
Note
If userApplication is force started on the operation node,cluster fail-over does not take effect even if userApplication is started on the standby node.Make sure that run the following command on the operation node to enable cluster fail-over.
# /opt/FJSVhanet/usr/sbin/hanetpoll on -f yes |
