Today we will look at the advantages of two types of switch aggregation: Switch Stacking, or switch stacks, and Chassis Aggregation, or switch chassis aggregation. This is section 1.6 of the ICND2 exam topic.
When designing a company network, you will need to consider the placement of Access Switches to which many user computers connect, and Distribution Switches to which these access switches connect.
The diagram shows the Cisco model for OSI Layer 3, where access switches are labeled A and distribution switches are labeled D. You can have hundreds of devices on each floor of a company building, so you will need to choose between two ways to organize switches.
Each of the Access level switches has 24 ports, and if you need 100 ports, then this is about 5 of these switches. Therefore, there are 2 ways: increase the number of small switches or use one large switch with hundreds of ports. The CCNA topic does not cover switch models for 100 ports, but you can get such a switch, it is quite possible. So, you have to decide what suits you best - several small ones or one large switch.
Each of the options has its own advantages. You can set up just 1 large switch instead of setting up several small ones, but there is also a drawback - there is only one point of connection to the network. If such a large switch fails, the entire network will collapse.
On the other hand, if you have five 24-port switches and one of them breaks, agree that the chance of one switch failing is much greater than the chance of all five devices failing at the same time, so the remaining 4 switches will continue to ensure the existence of the network . The disadvantage of this solution is the need to manage five different switches.
Our diagram shows 4 access switches connected to 3 distribution switches. According to layer 4 of the OSI model and Cisco network architecture requirements, each of these 2 switches must be connected to both distribution switches. When using the STP protocol, one of the XNUMX ports of each Access switch connected to the Distribution switch will be blocked. Technically, you will not be able to use the full bandwidth of the switch, because one of the two links is always down.
Usually all 4 switches are located on the same floor in a common rack - the photo shows 8 installed switches. There are a total of 192 ports in the rack. In this case, firstly, you must manually configure IP addresses for each of these switches, and secondly, configure VLANs everywhere, and this is a serious βheadacheβ for a network administrator.
There is a thing that can make your task easier - Switch Stack. In our case, this thing will try to combine all 8 switches into one logical switch.
In this case, one of the switches will play the role of the Master switch, or the owner of the stack. A network administrator can connect to this switch and perform all the necessary settings, which will automatically be distributed to all switches in the stack. After that, all 8 switches will work as one device.
Cisco uses various technologies to stack switches, in this case this external device is called the "FlexStack module". On the back panel of the switch there is a port where this module is inserted.
FlexStack has two ports where connecting cables are inserted: the bottom port of the first switch in the rack is connected to the top port of the second, the bottom port of the second to the top port of the third, and so on until the eighth switch, the bottom port of which is connected to the top port of the first switch. In fact, we have a ring connection of switches of one stack.
In this case, one of the switches is selected as the master (Master), and the rest - slaves (Slave). After using the FlexStack modules, all 4 switches of our circuit will act as 1 logical switch.
If the Master switch A1 fails, all other switches in the stack will stop working. But if the A3 switch breaks, the remaining three switches will continue to work as 1 logical switch.
In the initial scheme, we had 6 physical devices, but after organizing the Switch Stack, there were only 3 of them: 2 physical and 1 logical switch. In the first option, you would have to configure 6 different switches, which is already quite troublesome, so you can imagine how laborious it is to manually configure hundreds of switches. After stacking the switches, we got one logical access switch, which is connected to each of the distribution switches D1 and D2 by four communication lines combined into an EtherChannel. Since we have 3 devices, one EtherChannel will be blocked by the STP protocol to prevent traffic loops.
So, the advantage of a switch stack is the ability to manage one logical switch instead of several physical devices, which simplifies the network setup process.
There is another switch aggregation technology called Chassis Aggregation. The difference between these technologies is that the organization of the Switch Stack requires a special external hardware module that is inserted into the switch.
In the second case, several devices are simply combined on one common chassis, as a result of which you have a so-called aggregation switch chassis. In the photo you see a chassis for Cisco 6500 series switches. It combines 4 network cards with 24 ports each, so this unit has 96 ports.
If necessary, you can add more interface modules - network cards, and all of them will be controlled by one module - the supervisor, which is the "brain" of the entire chassis. This chassis has two supervisor modules in case one of them fails, which creates some redundancy, but improves the reliability of the network. Typically, such expensive chassis are used at the core level of the system. This chassis has two power supplies, each of which can be powered from a different power source, which also increases the reliability of the network in the event of a power outage at one of the power substations.
Let's return to our original scheme, where there is also an EtherChannel between D1 and D2. Typically, when organizing such a connection, Ethernet ports are used. When using a switch chassis, no external modules are needed; Ethernet ports are used directly to combine switches. You simply connect the first D1 interface module to the same D2 module, and the second D1 module to the second D2 module, and everything works together to form one logical Distribution Layer Switch.
If you look at the first version of the scheme, then to aggregate 4 access switches and a distribution suite, you need to use the Multi-chassis EtherChannel program, which organizes EtherChannel channels for each access switch. You can see that in this case there is a p2p connection - "point-to-point", excluding the formation of traffic loops, and in this case all available communication lines are involved, and we do not have a reduction in throughput.
Typically, Chassis Aggregation is used for high-performance switches, and not for less powerful access switches. The Cisco architecture allows the simultaneous use of both Chassis Aggregation and Switch Stack solutions.
In this case, one common logical distribution switch and one common logical access switch are formed. In our scheme, 8 EtherChannels will be created, which will work as one communication line, that is, as if we connected one distribution switch to one access switch with one cable. In this case, the "ports" of both devices will be in the forwarding state, and the network itself will operate at maximum performance, using the bandwidth of all 8 channels.
Thank you for staying with us. Do you like our articles? Want to see more interesting content? Support us by placing an order or recommending to friends, 30% discount for Habr users on a unique analogue of entry-level servers, which was invented by us for you: (available with RAID1 and RAID10, up to 24 cores and up to 40GB DDR4).
Dell R730xd 2 times cheaper? Only here in the Netherlands! Dell R420 - 2x E5-2430 2.2Ghz 6C 128GB DDR3 2x960GB SSD 1Gbps 100TB - from $99! Read about
Source: habr.com