HPE Community read-only access December 15, 2018
This is a maintenance upgrade. You will be able to read articles and posts, but not post or reply.
Hours:
Dec 15, 4:00 am to 10:00 am UTC
Dec 14, 10:00 pm CST to Dec 15, 4:00 am CST
Dec 14, 8:00 pm PST to Dec 15, 2:00 am PST
BladeSystem Virtual Connect
cancel
Showing results for 
Search instead for 
Did you mean: 

Virtual Connect LACP and NIC teaming Problem

 
chuckk281
Trusted Contributor

Virtual Connect LACP and NIC teaming Problem

Jay was having a customer issue:

 

************************

 

I have met an strange issue while I’m doing the VC test in customer site , Hope I can find some advice in here.

The test environment is like below:

 

 

And the configuration we set as below:

 

webpic72.png

 

1.All the Link Aggregation Protocol we use is LACP, All trunk protocol we use 802.1Q, and on Cisco Switch we set the EtherChannel mode to ”mode active”

2.We use NCU tool and set the BL460 teaming mode to “NFT”, two 460G7 belongs to two different VLAN

3.The server NIC use different vNet  and use an SUS on each VC modular to permit multiple VLAN,  use two channel group connect to 2960

4.Enable “SmartLink” setting on VC

 

The odd things is that:

1.All things work well and the Port Channel is formed but while we change the LACP mode on Cisco Switch, For example we change “mode active” to “mode passive”, The NCU will change the MAC address of two 460’s standby NIC, and will use 00-01-02-03-00-01 as the standby NIC’s MAC address

2.On Cisco 2960 switch we see the alert ”host 0001.0203.0001 in vlan 146 is flapping between Po3 and Po2”

 

Attached file is the screenshot for this issue, I want to know why  NCU change the MAC of standby NIC? And why use ” 00-01-02-03-00-01” as the MAC address

From the guide” Virtual Connect for the Cisco Network Administrator” , I know that VC can form channel while Cisco switch’s channel-group is set to either “mode active” or “mode passive”

 

Any suggestions is appreciated.  Thank you!

 

**************

 

Reply from Alok:

 

******************

 

I read the following in HP Proliant Network Adapter Teaming White paper.

 

When transmitting frames, the current Primary adapter always transmits using the team’s MAC address as the

Layer 2 address and the team’s Protocol address as the Layer 3 address. Non-Primary adapters always transmit

using the MAC address assigned to them by the teaming driver and using the team’s protocol address as the

Layer 3 address. For NFT and TLB, the MAC address used by Non-Primary adapters when transmitting is always

different from the Primary adapter’s MAC address and is always unique from that of any other Non-Primary

adapter, to comply with IEEE standards.

 

And also it is said that

 

If Cisco PVST+ is chosen under the Fast Path configuration section, the teaming driver configures the

teamed ports to listen for Cisco PVST+ BPDUs. This involves registering, in hardware, a specific multicast

MAC address on all teamed ports and monitoring for Spanning Tree frames in Cisco

PVST+ frame format. Since Cisco PVST+ is VLAN aware, configuring the VLAN for PVST+ to monitor may

be necessary. By default, Fast Path using Cisco PVST+ will monitor the first VLAN configured on the team.

The VLAN monitored by Fast Path can be changed on the VLAN configuration tab in NCU.

 

I assume the mac address “00-01-02-03-00-01” has something to do with the NIC teaming driver assigning the mac.

 

**********************

 

And from Jim:

 

******************

 

As indicated below with NFT, the MAC address used by Non-Primary adapters when transmitting is always different from the Primary adapter’s MAC address and is always unique from that of any other Non-Primary adapter, to comply with IEEE standards.

 

During the mode change are their factors that might contribute to the initiation of a failover event? What is the time interval between the mode change and the alert?

 

Spanning Tree’s blocking, listening, and learning stages should be disabled, or bypassed, on all switch ports to which the nic team port is attached. These stages are not needed when a non-switch networking device (for example, server) is attached to the switch port. PortFast which causes a Layer 2 LAN interface configured as an access port to enter the forwarding state immediately, bypassing the listening and learning states.

 

MAC flapping is when a source MAC address is seen on different ports. That could be an indication of a serious problem (like a loop) and the switch is letting you know about it. It can also cause excessive flooding and performance problems because the switch will be re-learning the MAC constantly every single time a frame is seen on a different port.

 

*******************

 

Reply from Jay:

 

******************

 

Thanks for your suggestion.

I think the PortFast setting  was not enable on the Cisco Switch, So I will change this setting later

And as you said  Non-Primary adapters always transmit using the MAC address assigned to them by the teaming driver and is always unique from that of any other Non-Primary adapter.

Is this means while the Non-Primary adapter transmit frames will use this unique MAC?  But why the Non-Primary adapter need to transmit frames? For the purpose of heartbeat?

 

*****************

 

Any other input for Jay?

1 REPLY
chuckk281
Trusted Contributor

Re: Virtual Connect LACP and NIC teaming Problem

Followup from Jim:

 

***************

 

All teamed ports may transmit and receive heartbeats, including Non-Primary adapters. A multicast MAC address is used in HP nic teaming for the purpose of transmitting and receiving heartbeat frames.

 

******************