Port Aggregation

Port aggregation is a technique to bind more than one physical port and create a bigger virtual port. It is done for multiple purposes.

The biggest advantage of this technique is increasing the bandwidth of a port. If you have two 100Mb ports and you need a port that can support 200Mb bandwidth then you just have to merge two ports and you will get the 200Mb bandwidth.

Another reason is redundancy if we want to provide resources with zero second downtime we can use this technique for redundancy and if one ports fails to work the other will be able to transmit the data and keep the services up and running.

This technique is quite popular and used for more power and efficiency. You can use up to 8 ports for aggregation purpose.

There is a question, Is this technology better than rapid spanning-tree in term of redundancy?

The answer is yes, because this gives more bandwidth and if the port goes down the other port still works where in spanning-tree only one port can work at one time.

Cisco uses three ways to apply this technology, Ether-channel, Port aggregation protocol (PAgP), and Link aggregation control protocol (LACP).

LACP is open standard of IEEE and it is used widely almost everywhere, if you have to configure aggregation go with it. PAgP and Ether-channel are Cisco proprietary and are not used now but you have to know all three.

Ether-channel is a static configuration mode and it is not quite used in production because if other switch fails to meet the requirements the port will no longer work. Where, PAgP, and LACP are dynamic and if switches do not match the configuration the ports work in normal way they do not stop working. It is recommended that you use dynamic technique and more specifically go with LACP.

Static Aggregation

Port Aggregation with Ether-Channel

We will work with the simple topology having two core switches.
It is a static method and is not recommended to use but it is there for use if you need to.

First we will get into CORE-1 shutdown the interfaces, and then configure the static aggregation mode. All we need to do is to make a group for channel this group is also called a port channel. It is important to keep in mind that you do the same configuration at both switches otherwise it could be difficult to troubleshoot if you get into an issue.

Interfaces are as following;

Spanning-tree status and neighbor information is as follows:

Configuration

Notice that channel-group supports only 48 groups, it can vary from switch to switch. Another thing to notice that there are 5 modes where 2 belongs to LACP and 2 to PAgP and only one is static which is on.

Now we need to do the same thing on CORE-2 as well.

Now we just need to turn on the switches ports and this is all.

The Po1 is the name of channel or we can say the name of virtual interface.
(SU) tells that channel is in Layer2 device and it is in used. The abbreviation for Flags has been listed in device to know the meaning of Flags.
Fa0/1(P) Fa0/2(P) shows that the two ports are part of port-channel.

Here are MAC addresses of Po1.

Look we have a new interface named Port-chanel1.

Notice that the interface has 300Mb of speed and it shows that Fa0/1, Fa0/2 are part of this channel.
Now if we see the status of spanning-tree.

Notice that there is only one interface which is Po1 and Type is Shr. There is no second link to the switch which could make a loop and we merged the two interfaces into one.

Now it is important that you deal these two ports at same time, if you make changes to one and do not do this for another then this channel will break. It is recommended that you make changes for Po1 interface so both the underlying ports are get treated equally.

Dynamic Aggregation

The dynamic configuration is not ideal in many cases but when it comes to aggregation, this mode is quite satisfactory. Because if you do static aggregation if there is any issue with aggregation the frames will not be transmitted but if it happens in dynamic mode the ports will get back to their normal behavior and frames will work fine.

There are two dynamic modes as we have discussed LACP and PAgP. The LACP is only better because it is open standard by IEEE and supported by all vendors, excluding this difference there is no underlying technical difference between these.
When configuring dynamic aggregation go with LACP because it is open standard and supported by all vendors.

Dynamic Modes

Active belongs to LACP and it means that the channel will continuously ask its neighbor to get aggregated if neighbor supports aggregation.
Desirable belongs to PAgP and it works same as Active, meaning that it keeps asking its neighbor to get aggregated.
Auto also belongs to PAgP but it is a bit shy one it does not ask to get aggregated but it waits for another party to offer invitation to get aggregated.
Passive belongs to LACP and it works same as Auto.

Aggregation LACP

Switch-1

Switch-2

Aggregation Result

Active

Active

Aggregation Successful

Passive

Active

Aggregation Successful

Active

Passive

Aggregation Successful

Passive

Passive

Aggregation Unsuccessful

 

 

 

 

Aggregation PAgP

Switch-1

Switch-2

Aggregation Result

Auto

Desirable

Aggregation Successful

Desirable

Auto

Aggregation Successful

Desirable

Desirable

Aggregation Successful

Auto

Auto

Aggregation Unsuccessful

 

 

 

 

LACP Aggregation Configuration

The steps are same as we did for static aggregation, if you have already a static aggregated channel you need to delete it first.

Now do the same for CORE-1 Switch.

Now turn on the ports.

Look now the ether-channel is using LACP mode which is dynamic.

There is a question, If there are both Gigabits and Megabits ports and you aggregate them, will they become aggregated?

The answer is Yes, they will become aggregated but the switch will only use Gigabit ports and if they fails to work then the Megabit ports will be used.
For example if you have aggregated 4 ports in which 2 are Gigabits and 2 are Megabits then only Gigabits ports will be used and Megabits will sit idle, even the bandwidth of Megabits ports will not be aggregated unless the Gigabits ports stop working.

Stackwise is another technique to get redundancy and more bandwidth. Read about it.

Prerequisites for 200-301

200-301 is a single exam, consisting of about 120 questions. It covers a wide range of topics, such as routing and switching, security, wireless networking, and even some programming concepts. As with other Cisco certifications, you can take it at any of the Pearson VUE certification centers.

The recommended training program that can be taken at a Cisco academy is called Implementing and Administering Cisco Solutions (CCNA). The successful completion of a training course will get you a training badge.

Full Version 200-301 Dumps

Try 200-301 Dumps Demo