Virtual Connect 4.10 - VMware 5.5 U1 Network Failover Problem
Hi,
yesterday i installed following System for my customer:
Hardware:
- 1x c7000 Blade Enclosure
- 2x HP Flex-10/10D VC Modules
- HP BL460c Gen8 with FlexFabric 10GB 2-Port Adapter
- 2x HP 5920 10GB Switches in an IRF Stack
Firmware:
- OA: 4.11
- VC: 4.10
- HP BL460c Gen8 Server is up-to-date (bootet with actual Proliant Service Pack DVD)
Configuration (Virtual Connect Active/Active):
- 20GB/s LACP Trunk configured on VC Modul in Bay 1 (SUS with Name "SUS-Bay1")
- 20GB/s LACP Trunk configured on VC Modul in Bay 2 (SUS with Name "SUS-Bay2")
- One LACP Trunk configured on the IRF Switch for Bay 1 and one for Bay 2
- SmartLink is activated for each Network
Server Profile:
- Port 1 configured with Network on "SUS-Bay1"
- Port 2 configured with Network on "SUS-Bay2"
Configuration (VMware ESXi 5.5 U1):
The ESXi Host has configured one vSwitch (vSwitch0) with "vmnic0" and "vmnic1". Management Network is on that vSwitch and uses "vmnic0" as Active Adapter and "vmnic1" as Standby Adapter. In the Network Settings in leaved everything on default Settings.
Please see attached Screenshot.
Problem:
After installing the c7000, VC and one VMware ESXi Host i wanted to try some Failover Scenarios. So i disconnected all FC Cables from VC Modul in Bay 1 and pinged the Management IP from the ESXi Host at the same time. "vmnic0" was disconnected and i was loosing one ping. Thats fine. Management now on "vmnic1".
After that, i reconnected the FC Cables back to VC Bay1. Now, i was loosing round about 6-7 Pings (about 30 Seconds).
I (and my customer) were a little shocked because 30 Seconds can be really long...
On other VMware Installations this "failover" should generate maybe one lost of ping, not more.
So i need some help how to solve this issue. Anyone with the same configuration? Any Tips?
