We have a Routerboard [1100AHx2] with 6.32.2 router OS … and 3 HP managed switches connected to this routerboard, this form a bridge for our internal LAN. We have 2 VMWare 6.0 host with 3 NIC each, they are teaming & each of them is connected to a HP managed switch. On each host has 3 VM servers, let say host 1: A-B-C & host 2: D-E-F … these servers are logically traverse via the routerboard on different HP switch… atm, A via Sw2, B&C via Sw1, D&E via Sw3, F via Sw2.
With this topology we experience some strange behavior that:
- Client PC from Sw1 could not ping server D & F… but can ping the rest 4.
- Client PC from Sw2 could not ping server D… but can ping the rest 5.
- Client PC from Sw3 could not ping server F… but can ping the rest 5.
- Client PC from each Sw, ping to other clients of other Sw: most of them success, some of them fail…
It’s hard to understand that both server D & E are logically connected to Sw3… but client from Sw1 & Sw2 could ping E, not D… same for server F & A are logically connected to Sw2… but client from Sw1 & Sw3 could ping A, not F…
Some facts:
- VMWare host use NIC teaming with default load balance option [ original port base ].
- No VLAN implementation.
Please help to advise while pinging within a bridge cause inconsistent behavior ?!
Thanks,