Transparent HA fail-over issue

[Jeff_Gao]

Dear

      I encountered a HA fail-over problem in transparent HA mode.Pls see the below infomations:

Hardware module:5800

Software version:R80.30+HF215

1.WAN Access SW is a layer2 switch and stack;

2.checkpoint firewall is transparent mode,compose of bridge1 with bond1 and bond2 ,bond1 and bond2 is lacp mode;bond3 is sync and the mode is active-backup;

3.Core SW is Core switch,

The issue is HA fail-over when use clusterXL_admin down/up.Network interruption when i execute clusterXL_admin down;clusterXL_admin up in FW01(Active).

According to my demo test,when execute cli fail-over,switch mac-address table and arp table no refresh,so network traffic interruption.when i manual refresh switch arp table,all work.For example:

when active is FW01,core switch to WAN router(CUC MPLS router or CTC MPLS router)  mac-address and arp are recording to port Eth-trunk1,I execute fail-over cli in FW01 and switch active to FW02,but switch still recording mac-address and arp on Eth-trunk1 until arp table timeout,and network interruption during.

I asked the switch vendor,they said that switch can not perceive firewall fail-over,so mac-address and arp can not be refreshed.

Now,we exclude the switch,how to trigger switch refresh mac-address table and arp table when firewall fail-over with cli.

thanks!

[PhoneBoy]

Is Spanning Tree in use in this environment?

[Jeff_Gao]

yes,stp in use. But i also try to disable stp,not work. I understand that standy firewall should drop all traffic include bpdu,right?