Need to shut down 1 link in 40 Gbps LACP Bundle

[Dean Romanelli]

Hi All,

Hopefully an easy one: I have a 4 port, 40Gbps port-channel between two of my data centers for 6509 VSS, connected with fiber. One of them is dirty --> incrementing about 100 CRC errors every 3 seconds.  Obviously this is causing problems, so I need to shut this 1 link in the bundle down, as 30 Gbps is still plenty of bandwidth.

Will this affect services? If there is traffic using that link at the time of shutdown, will it get retransmitted?

[Ganesh Hariharan · Accepted solution]

Hello Dean,

As per my exp. If you configure EtherChannel load balancing on a switch that is not in VSS mode, traffic is disrupted while the EtherChannel member ports transition through the shutdown and then no shutdown states and There would be no disruption on switches in VSS mode.

Also depends on type of traffic flowing between DC's Certain application are very sensitive that they cant have take a single breakge in TCP communication between Apps abd DB servers, But most of the TCP based application will re-tranmit to reagin connection.

Hope it Helps..

-GI

Rate if it Helps...

[Dean Romanelli]

Thank you Ganesh.

Now here's the other thing, the 10 gig links are on the supervisor engine cards.  If I shut one of the 10 gig links down, I just want to make sure the supervisor engine will not try to fail over.

[Reza Sharifi]

Hi,

Disconnecting the 10Gig links will not cause Sup failover. Only removing the primary sup from the chassis will cause failover or if you force failover using IOS. 

If your Portchannel is configured correctly and you disconnect one of the links, you will only lose what is already in the wire and if that interface is not being used at the time of disconnect than you don't lose any packets.

HTH

[Dean Romanelli]

Thank you all very much.

[Masoud Pourshabanian]

Hello,

Please take a look at the link below.

"If a single VSL physical link goes down, the VSS adjusts the port group so that the failed link is not selected.

If the VSS standby chassis detects complete VSL link failure, it initiates a stateful switchover (SSO). If the VSS active chassis has failed (causing the VSL links to go down), the scenario is chassis failure, as described in the previous section.

If only the VSL has failed and the VSS active chassis is still operational, this is a dual-active scenario. The VSS detects that both chassis are operating in VSS active mode and performs recovery action. See the "Dual-Active Detection" section for additional details about the dual-active scenario."

http://www.cisco.com/c/en/us/td/docs/switches/lan/catalyst6500/ios/12-2SX/configuration/guide/book/vss.html

Masoud