Etherchannel is a Cisco term that allows a selection of ports to be grouped into a single virtual connection.

PROS

1. Able to move more data than a single dedicated line.
2. Provides Redundancy & Instantaneous failover if a line/port does down.
3. The switch views the virtual connect essentially just like any other line. ie Protocols like STP and VTP work over it.

CONS

1. More Ports used on the switch may have a great cost associated with it.
2. 4x100Mb/sec lines means there is still only a max of 100Mb/sec for each individual conversation.
3. All Ports have to be Identical (Ie. All FastEthernet or All Gigabit)
4. All ports have to be configured the same (ie. Duplex, Speed etc)

Setting up Etherchannel

There are 3 main steps when setting up Etherchannel

1. Specify the ports needed
2. Specify Etherchannel Protocol. There are 2 main protocols:
   • a.) IEEE – LACP
   • b.) Cisco – PAgP
3. Specify the Channel Group ID and Mode. This is done in the same command.
   • a.) Because you can have multiple Etherchannels per switch you will need to specify an ID for each one
   • b.) Depending on if you choose LACP or PAgP you will need to specify different modes.
4. Turn the Ports into trunking mode
5. Because Ports are in trunking mode please also turn on encapsulation ie. dot1q

Commands to setup Etherchannel

Here’s a screenshot from Live Cisco 2960 Switches being setup with Etherchannel:

Switch1

Switch2

What do the happens when something goes wrong with one of the lines with the etherchannel?

Switch2 – Unplug Port fa0/22, wait a few minutes and then plug the port back in:

Let’s go over the last image.

1. Unplug cable.
2. Look at how fast the etherchannel reports the issue. It’s actually BEFORE the port actually goes down!
3. Port goes down
4. I hit <enter> a couple of times to break up the output.
5. Plug the cable back into port fa0/22
6. Interface comes back up
7. Etherchannel rebuilds itself

Main commands to troubleshoot:

show etherchannel summary

• Displays info on status of group, port-channel, protocol and port

show etherchannel port-channel

• Displays configured properties of port-channel, port state, protocol, port security, fast switchover and load share. It also lists the ports and timing info.

show interface trunk

• Displays info about all the trunk interfaces.

Are you getting the error: Command rejected: An interface whose trunk encapsulation is “Auto” can not be configured to “trunk” mode?

I first got this error when I was trying to setup a Cisco 3560. To be more precise

Well this is going to be a short post you might see on a Cisco router. The error I will be going over today is the “error” – “Can’t use same dlci on overlapping circuits”.

There is a few reasons why you are getting this error:

1. You’ve already entered that DLCI on the router somewhere or
2. You’ve already entered that DLCI on the router somewhere

Lol.

Your best friend is to do a “show run | include int” command.

This will show you all the interfaces neatly on the screen. I’m willing to be you will find that DLCI has already been configured on the the router by you or someone else.

This has happened to me a few times because I’ve saved my lab after I set up the DLCI by accident. Then when I visited the lab a week or two later I got this weird error.

Now if you ever get that “Error” message “Can’t use same dlci on overlapping circuits” you will know what to do!

Frame Relay:

• Gives the service provider the ability to make a “cloud of bandwidth”
• Service providers realize that no one uses 100% of their bandwidth 100% of the time. Because of this they can over commit the bandwidth.
• Make it cheaper to get connections over point-2-point leased lines.

Point-2-Point Looks like this:

Frame Relay kind of looks like this:

Frame Relay Terminology you will need to know

Committed Information Rate (CIR)

• Logical speed of the line.
• Min Bandwidth the service provider will give you

Local Access Rate (LAR)

• Speed of the Physical cable. It will most likely be > than the CIR
• Physical speed of the of the cable you connect your equipment to the provider with.

Local Managment Interface (LMI)

• The Language you speak between your router and service provider (3 types… Cisco, ANSI and one other)
• Signaling Protocol. Provides statistics from the service provider on the line
• Status – Quality and dropping of packets
• LMI can send DLCI info.
• NOTE – See diagram above for LMI

Data Link Connection Identifier (DLCI)

• Ethernet uses MAC addresses, Frame Relay uses DLCI
• DLCIs are LOCALLY SIGNIFICANT ONLY!!!!!!!!

Permanent Virtual Circuit (PVC)

• It is the logical circuit between routers in frame relay
• This is something the service provider sets up.
• Every PVC has a CIR
• The Greater # of PVCs, the Grater the Cost ($)

How DLCIs work

DLCIs can have a number from 16 to 1024.

I watched a presentation that explained DLCs and I have to say it was the best description I have ever heard it went something like this.

I head to the New York Airport because I need to take a flight to Toronto. When I get to the airport I look at my ticket and see that in order to get to Toronto I need to go to “New York Gate 102”  to get on the plane. I then get on a plane and trust that the pilot will get me safely to Toronto. When the plane lands safely in Toronto we get off the plan and come through “Toronto Gate 201”. I then leave the airport and call a taxi to where I need to go.

The Gates are like DLCI’s. Unlike in Ethernet where the MAC address is the source address on your local computer that eventually talks to a remote MAC address on another, the DLCI closest to where you leave is actually the Destination Address and the DLCI where the packet leaves the frame relay cloud is actually the address of where you came from if you took the trip backwards!

So while DLCIs are like MAC in that they are addresses for Frame relay they are interpreted much, much differently. Basically in reverse!

So DLCI 607 is actually the destination address of Router 7. DLCI 706 is the destination address of R6 FROM R7. The two DLCI’s on Router R6 can NEVER be the same because it is LOCALLY SIGNIFICANT. That being said R7 and R5 could both be changed to 607 and R6 wouldn’t care as long as the Frame Relay cloud was configured properly by the service provider.

Let’s look at it one more time! 605 is the destination of R5 on R6. On R5 506 is the destination address of Router R6! It’s important to understand DLCIs!!!!!

One more thing you should note is that the red lines represent the PVC between the routers!

Frame Relay PVC Design

There are 3 ways you can design the PVCs. They are

Hub & Spoke

• Cheapest to deploy
• Single Point of failure
• Delays can become an issue

Full Mesh

• Exponentially Expensive! Each time you add a new location, you then need to create redundant PVCs to all other routers as well.
• Because of this it doesn’t scale well.

Partial Mesh

• Best compromise on performance and price
• There is redundancy on the most important parts of the network only where needed.

Frame Relay Point Design

Besides PVCs you can also design how each point is designed.

Multi-Point Design

• Each router is in the same subnet.
• Multiple DLCIs are mapped to an interface
• Causes problems with Split Horizon. ie. Typically a router will not allow routes it learned to leave on the same interface. This prevents routing loops.

Point-2-Point Design

• All Routers are on different subnets
• Point-2-Point sub-interface created for EACH peer
• Split Horizon isn’t an issue for P-2-P design.

I hope this helps you with the basics of frame relay!

Here’s some of the commands I type to get GNS3 working a little bit better. By specifying the ports to use a specific speed and duplex you can avoid the dreaded Mismatch and the Module not working issue.

Please note in GNS3 I stictly use Cisco 2691 routers. When needed I add the “NM-16ESW” adapter to slot1.

NOTE – If you are creating a project file it’s best to save the config the way GNS3 likes. To do this configure the device to the base config needed for the lab.

• Right click the device – Select “Startup-config”
• Click “Load from nvram”
• Hit “Apply” and then “OK”
• The Device if rebooted will now load this base config.

PC

Typical Command for PCs.

• conf t
• int fa0/0
• speed 100
• duplex full
• ipv6 address autoconfig
• no shut
• end
• wr
• !

Router

• conf t
• int range fa0/0 – 1
• speed 100
• duplex full
• end
• wr
• !

Router with Switch (NM-16ESW)

• conf t
• int range fa1/0 – 15
• switchport
• speed 100
• duplex full
• end
• wr
• !