NGX R61 Adding New interface drops Cluster

[Kurgen727]

NGX 61 2 server (FW3 active , FW4 standby) in an active/standby setup with 8 interfaces Eth0-7.

I previously had eth0-4 configured. Network Interface for eth5 is not yet connected to the switch.

I went into sysconfig on FW4 (standby) to add a connection, config the adapter, to eth5 and poof the whole cluster went offiline. I tried stopping and restarting the member but no luck. I deleted the connection, restarted both gateways and the cluster was restored.

Later that same evening I repeated the procedure, and everything went flawlessly.

-The next day I connected the eth5 interface (only after I ran the command "ifconfig eth5 down") to the switch, and the cluster again crashed.

I finished the cluster config in dashboard, for the new adapters and everything was fine again.

Should adding an adapter (on the standby no less) take down the whole cluster? Is there a way to avoid this? I've never had this happen before.

Thanks.

[Routerkid1]

Well here are my rules to go by.

1. Do not configure an interface without plugging it in to the switch as Cluster XL will send ccp packets toward this interface and this can cause a interface clapping issue. You can check the control connections in tracker and see if you see any that show recieve up transmit down.

2. If you must configure the interface and do not have the switch in place mark the interface as monitored private and reinstall the policy.