Etherchannel help

Hello all!

This might be a bit of a newbie question, but I always had a hard time with etherchannel and HSRP/VRRP/GLBP.

If I only set
(conf-if)#channel-group 1

on each of the interfaces I want in the Po and don't stipulate channel mode, what will happen?

Find more posts tagged with

Free for TechExams community: Cybersecurity salary guide

Compare cert salaries and plan your next career move

Button

Comments

ColbyG

Lab it.

Netwurk

Lab it might sound like a "no help" answer, but Colby has a point. For CCNP level stuff, you really need some hands on.

If you don't have your own equipment (or access at work), consider renting some rack time.

DevilWAH

well if you read the documentation and course material you should be able to answer that question easily.

but even so you want to lab it to check you have understood it right.

I was trained as a scientist and there method is. Form a idea, test that idea and come to a conclusion. However this is by far the best method to learn.

what do you think will happen and why??

now go test it. If it happens then you understand it, is some thing else happens then try to figure that out and retest...

mikej412

creamy_stew wrote: »

what will happen?

The defaults for the hardware/ports

creamy_stew

DevilWAH wrote: »

what do you think will happen and why??

Well, I can't say I've read and digested the documentation for all cisco Catalysts. Nor have I read the course material (I haven't taken the official course)

My initial thought was that PAgP was the default, and so setting the channel mode would be unnecessary. However, When going through the practise exam that came with the BCMSN cert guide, they said to set it to "on" implying that neither LACP nor PAgP are enabled by default.

I do have access to switches at work, however, I don't don't like to use production equipment as a lab. I had a 3750 and 3 2960's at home when I studied the most. Alas, they are now deployed.

Oh well, hopefully I can scramble 2 2960's to try out some more etherchannel configs.

BTW, some of you seem to think that this behavior is somthing that is set in stone and that everyone should know. I'd say that this is something that is very likely to change between models and IOS versions.

edit: Heh, I came off looking a bit defensive, didn't I

Obviously you all think this is well worth labbing out. My practise exam suggests I should put more effort into this also. (66% on par with Wireless, which I haven't even studied yet)

DevilWAH

creamy_stew wrote: »

My initial thought was that PAgP was the default, and so setting the channel mode would be unnecessary. However, When going through the practise exam that came with the BCMSN cert guide, they said to set it to "on" implying that neither LACP nor PAgP are enabled by default.

Well sounds good to me your reasoning.
Most people recomend you set etherchannels to ON, and I agree if you are running switch to switch channels. When setting up switch to server it can be more flexable to use protocol methods of configuring. But in generaly forcing On means one less thing to worray about going wrong.

BTW, some of you seem to think that this behavior is somthing that is set in stone and that everyone should know. I'd say that this is something that is very likely to change between models and IOS versions.

Actuly CISCO keep things pretty constant across IOS and versions, I have had very few issues when upgrading switches and IOS where the defaults have caugh me out.

Netwurk

Back to the original question (although I still think he should lab it

)

creamy_stew wrote: »

If I only set
(conf-if)#channel-group 1

on each of the interfaces I want in the Po and don't stipulate channel mode, what will happen?

If you just enter channel-group 1, you'll get

% Incomplete command

You need to specify a mode. If you specify "on", you'll be fine if the switch on the other end is also set to channel-group 1 mode on

On an older IOS switch (a 2900 or 3500), the command would be port group 1

ColbyG

creamy_stew wrote: »

BTW, some of you seem to think that this behavior is somthing that is set in stone and that everyone should know. I'd say that this is something that is very likely to change between models and IOS versions.

Incomplete commands are pretty consistent across different hardware and code versions.

Netwurk

ColbyG wrote: »

Incomplete commands are pretty consistent across different hardware and code versions.

Yes, that's true now. But prior to IOS 12.2, there was a lot that varied between the various models and OSes.

You had the CatOS commands, the 19xx series commands, and the IOS 12.0 commands which have quite a few differences compared to 12.1 and 12.2

Anyhow, labbing up switches with the modern commands should be sufficient as CatOS, 19xx, etc. is no longer covered on the exams

ColbyG

Netwurk wrote: »

Yes, that's true now. But prior to IOS 12.2, there was a lot that varied between the various models and OSes.

You had the CatOS commands, the 19xx series commands, and the IOS 12.0 commands which have quite a few differences compared to 12.1 and 12.2

Anyhow, labbing up switches with the modern commands should be sufficient as CatOS, 19xx, etc. is no longer covered on the exams

It seems you missed the joke.

Netwurk

ColbyG wrote: »

It seems you missed the joke.

I never joke about switches unless they are 1900's

Stotic

DevilWAH wrote: »

Actuly CISCO keep things pretty constant across IOS and versions, I have had very few issues when upgrading switches and IOS where the defaults have caugh me out.

There are intricacies with port-channels I've found out in the real world when aggregating ports on different card types. I forgot the issue, but I'm sure you'll run into it one late night lol

ColbyG

Netwurk wrote: »

I never joke about switches unless they are 1900's

We have some 1900 switches and 1600 routers in storage here. I chuckle when I see them.

sides14

DevilWAH wrote: »

I was trained as a scientist and there method is.

Good going in squeezing in the Scientific Method. Next will be the six step process for troubleshooting.

Definitely a method to the madness.

creamy_stew

ColbyG wrote: »

Incomplete commands are pretty consistent across different hardware and code versions.

Oh, God. Now I get why you were so quick to say to lab it out

On the plus side, I don't think I'll make any mistakes on channel mode on the exam now, and all it took was a little bit of embarrassment.

Etherchannel still worries me, though. I find it hard to predict how the different lb schemes will impact traffic patterns and why one wouldn't always use src/dst ip + ports. Is this not available on lower end IOS's/platforms? (sp?)

edit: 100th post! I know I shouldn't, but I'm going to celebrate by having "a" beer.