Spanning Tree Algorithm (spa)

To know about STP (Spanning Tree Protocol) and also known as Spanning Tree Algorithm (SPA). It is a layer 2 protocol which usually runs on all the switches or bridges and also its specification is that, it is an IEEE 802.1 D. as we know that sometimes loops are created in the network or while designing a network. A loop can be a bridge loop or a switching loop and is between two end points or two ports. This loop creates a storm of broadcasts and repeatedly broadcasts the same messages and hence flooding the network with the same messages. So, the main task or purpose of STP is to just not let the loops get created in the network and instead use the paths that are not used or the redundant paths in the network and not let the loops get created, thus avoiding the flooding of the network. So, these redundant links are important as whenever a failover occurs, there are these links available, so the users can use the network without any interruption.

In a network, all the switches must select a root switch and so that the entire network can be configures as per that root switch. But, in a case of VLAN’s they must have its own root bridge as each VLAN is a separate network or a broadcast domain. While selecting the root switch we have to be very careful, we can select the root switch or let the switches select the root, but the secondary can be harmful or risky as some suboptimal paths can be left in the network. Now, the switches for the selection of the switches sends some information to each other that is most important or plays an important role in the root selection. All this information is sent with the help of BPDU’s that is Bridge Protocol Data Units. Now, it is quite simple in work. The information from two switches is compared and the one with the good information is kept and the others is removed or neglected. Now, when it comes to select Root Bridge and then they send the frame called BPDU contains of the Bridge priority and the MAC address.

Bridge ID = Bridge Priority + MAC address

So, when the switch sends the information, so the bridge ID of the root bridge and the bridge is of the transmitting bridge is the same, as they claim to be the Root Bridge. And then all this is compared to select one Root Bridge. And so the root bridge is selected based on the lowest bridge ID. But, if the bridge ID of two switches is the same, then it is compared based on the MAC address. And, the user can select which switch will be the Root Bridge, by simply lowering the Bridge ID. Because once the bridge ID value is lowered that switch will become the Root Bridge and thus we can operate the entire system as our suitability and can avoid the unnecessary harm that will be caused by the system if we leave the root selection on the system.

Root port: The root port is the port that is the closest to the root bridge and the main and the most major thing is that one switch can have only one root port, the reason being that the information that is being transferred from the switch to the root bridge is done by only one port and so again to avoid the confusion and also the errors that might occur if there are two ports assigned to only one switch. But, the root bridge is the one that cannot have a root port as we know that the root port is used to point the information or direct the information to the root bridge, all this is then determined by the path cost. The path cost is the cumulative path cost to the root bridge. The path cost is the one that is dependent on the bandwidth.

So, the higher the bandwidth the lesser is the path cost and so as the bandwidth decreases, the path cost increases. To know more in detail, the messages that are sent by the switches to the root bridge are actually sent by the path that has the lowest cost, that means that the switches automatically select the path with higher bandwidth and that is because of the higher bandwidth the messages can transfer faster to the root bridge without or with minimal loss of message. All the switches in the network have the path cost but the root bridge is the one that does not have the path cost and that is the root bridge is the one which has the path cost as ‘0’. Whenever a switch sends the message to the other, it always adds the cost and then sends the message to the next switch. But, when it comes to the root bridge, it does not have a path cost. So whenever the message is sent from the root bridge to the other switches, the path cost from the root bridge is sent as ‘0’ and so whenever the switch adds its own path cost is added to it. So, when the root bridge sends out the messages it shows path cost as ‘0’. Here, there are switches which consist of the low path cost and also the switches that have comparatively higher path cost. In these the switch which has the lowest path cost is considered superior switches as compared to other and the switches that has the higher path costs is considered as inferior BPDU.

In addition to all these, STP selects one port on only one particular segment to send or instead forward traffic and on the ports on which the messages are mot sent are told as undersigned ports and are blocked. But, in this the ports only cannot send the data but can receive the BPDU from other switches. After all this is done, the network comes to a stage called convergence. Convergence occurs when all the ports in the STP reaches or have transitioned to one of the two stages and they are either the blocking or the forwarding state. Now, this convergence has to be done fast and this stare in STP is very important and the time of convergence should also be less. In large networks, fast convergence is even more desirable. The normal time is about 50 seconds for the switches that we have used i.e 802.1 D and this timer can also be adjusted or controlled by the user.

The entire working of this happens in the following way. When in a network the STP is enabled on the switch, the switches in the network goes through the process of blocking and forwarding state and once the state is selected, it then decides the state and then they get stabilized on their respective states. In the blocking state no data or messages is transitioned or transferred or either sent or received but the BPDU data is received and in which also the frames are discarded. But, there is also an exception in this state, if the port that is being used as the forwarding port is failed or disabled, then the other port which is blocked is then enabled to transfer the data or the messages in the network and also this can be done by the user of the network intentionally for some use in the network. The second state is the listening state. In this state the switch takes the BPDU message and waits for the information that can block the port or can make the enabled port for forwarding as blocked port and this state also discards the frames. The third state is the learning state, in which the BPDU’s are either sent or received and also it reads the MAC addresses, but this state still discards the frames. The last sate is the forwarding state, this is a normal state which sends, receives the data and also the BPDU data and reads the MAC addresses.

In this experiment we have connected in total switches and made one of the switch as root switch and then after making all the connections we found out the cost for each path, so that the messages can be sent through the lowest path cost to the root bridge. In this network one of the switches has a port which is blocked and so we get to know the actual path cost that the data has taken to reach the root bridge and also known about the commands to check the blockedports, the root bridge and also the brief description of the switches. Over here all the switches have the same path cost and that is 19, but if the other paths would have the smaller path costs, then we could have known that the switch can take the path that has the lowest cost.