2011-03-10 11:31 AM
The networking team will be performing some network maintenace which will affect connectivity to 2 (clustered) NetApp heads. Is there anything else other than cf disable that I need to do in order for the cluster not to fail?
2011-03-10 11:51 AM
Unless you have both "options cf.takeover.on_network_interface_failure on" and "nfo" on the ifconfig of any interfaces, the network drop will not cause a failover. You can check the options and also ifconfig -a to see if nfo is applied to any interface. By default, network failover is not configured in options and also has to be enabled in /etc/rc on the ifconfig lines. If you are running nfo, you can "ifconfig interface -nfo" on each interface to remove nfo, then "ifconfig interface nfo" to re-enable after maintenance. This might be a better option than cf disable since cluster failover would still be available for any other non-network hardware failures during the maintenance
2011-03-10 01:45 PM
I checked the rc config and did not see any settings related to "nfo." Here's the output of the 'options cf.takeover.on_network_interface_failure' command. It looks like it is disabled.
NetApp> cf.takeover.on_network_interface_failure off
What would be the best approach for this scenario?