Cisco Support Community
cancel
Showing results for 
Search instead for 
Did you mean: 
Announcements

Welcome to Cisco Support Community. We would love to have your feedback.

For an introduction to the new site, click here. And see here for current known issues.

New Member

etherchannel

hi! I'm trying to configure etherchannel between my blade server's cisco sw module with my old cat6509 sw. The sw port is currently running 10/100 mbps and my blade server's sw module is running on 1gbps. Is it possible to create 2 channels of etherchannel links in this case? (one etherchannel link to core1 and the other to core2). Can this be done with the different port speeds between the core and the server' sw? will there be any looping issue?

Thanks.

1 ACCEPTED SOLUTION

Accepted Solutions
Cisco Employee

Re: etherchannel

Hi David,

As long as the ports are capable of autonegotiation, you do not have to set the speed or duplex statically. I was asking, though, because I have had an experience with a Dell blade switching module that was fixed at 1Gbps/full duplex and was not able to work with any other combination of port speeds and duplexes.

How to check whether a loop has formed? First, I would go after the "show spanning-tree" output on all switches that are connected in a physical loop. Check all the ports that connect the switches together and whose sequence forms a physical loop. At least one of these interconnecting ports among the switches must in Blocking (or Discarding) state. If it is not, you most probably already have a Layer2 loop. An existence of a loop will always have a detrimental effect on the performance of the network and devices but the actual symptoms may vary wildly, from LEDs just blinking crazily and feeling that the network "just does not feel well" up to a network crash and loss of entire connectivity. In any case, if a frame starts looping in a network, you will notice an exorbitant number of received/sent frames on an interface (using MRTG/RRD to gather average load data for individual ports can help here significantly), potentially an increased load on the CPU and various other indications of an unusually high load in the network.

That being said, the STP protocol, when deployed properly (which, ironically, is often deployed properly if it is not overconfigured), will prevent your network from creating Layer2 loops.

Best regards,

Peter

3 REPLIES
Cisco Employee

Re: etherchannel

Hello David,

Is the blade switch capable of running 100Mbps on its ports? If it is a fixed 1Gbps switch then the ports would not even link with your Cat6500. The link would not even come up.

However, if the blade switch is capable of running 100Mbps then there will be no problems if the Etherchannel is configured properly. Pay attention to identically configure all member ports on a switch before grouping them under a Port-channel interface.

Regarding loops, you should definitely run STP or RSTP in your network. A Port-channel is treated as a single interface by the STP. If the Core1 and Core2 are interconnected then you indeed have a physical loop and you will need STP to prevent creation of a Layer2 loop.

I am not sure if this answers your question so please feel free to ask further.

Best regards,

Peter

New Member

Re: etherchannel

hi! I believe the blade's sw module should be able to run on 100mbps. Does that mean that i need to fix the speed to 100mbps at the blade side?

Normally, how do you know/check on the system whether a loop has actually formed?

Thx

Cisco Employee

Re: etherchannel

Hi David,

As long as the ports are capable of autonegotiation, you do not have to set the speed or duplex statically. I was asking, though, because I have had an experience with a Dell blade switching module that was fixed at 1Gbps/full duplex and was not able to work with any other combination of port speeds and duplexes.

How to check whether a loop has formed? First, I would go after the "show spanning-tree" output on all switches that are connected in a physical loop. Check all the ports that connect the switches together and whose sequence forms a physical loop. At least one of these interconnecting ports among the switches must in Blocking (or Discarding) state. If it is not, you most probably already have a Layer2 loop. An existence of a loop will always have a detrimental effect on the performance of the network and devices but the actual symptoms may vary wildly, from LEDs just blinking crazily and feeling that the network "just does not feel well" up to a network crash and loss of entire connectivity. In any case, if a frame starts looping in a network, you will notice an exorbitant number of received/sent frames on an interface (using MRTG/RRD to gather average load data for individual ports can help here significantly), potentially an increased load on the CPU and various other indications of an unusually high load in the network.

That being said, the STP protocol, when deployed properly (which, ironically, is often deployed properly if it is not overconfigured), will prevent your network from creating Layer2 loops.

Best regards,

Peter

143
Views
4
Helpful
3
Replies