Ethernet has become one of the most widely accepted protocols, and we will dedicate some time to discussing it. In the field of networking, there are various types of redundancy tailored to specific applications, such as equipment redundancy and connection redundancy. For this session, we will concentrate on two primary categories of redundancy
Path Redundancy: This type of redundancy ensures availability through multiple paths to a specific destination.
Link Redundancy: This redundancy involves having more than one port or physical connection to a destination.
While link and path redundancy are beneficial for network availability, they can also lead to a potentially devastating issue known as an Ethernet switching loop. I will elaborate on this in the next session.
This occurs when there is more than one Layer 2 path between two or more endpoints. To populate MAC tables, a typical switch broadcasts the first time it attempts to reach a new MAC address. When multiple paths exist, the broadcast is sent along all paths except the originating one. The other endpoint issues a similar broadcast, and the frame is continuously rebroadcasted, causing a switching loop. Since Layer 2 frames lack a TTL (Time to Live), they can remain in the network indefinitely, and often the resolution involves powering down the affected switches.
Nokia, as well as the broader industry, has implemented two main solutions to address this issue:
In the SROS, STP and priorities are configured inside of a service and not in the base router.
This occurs when there is more than one Layer 2 path between two or more endpoints. To populate MAC tables, a typical switch broadcasts the first time it attempts to reach a new MAC address. When multiple paths exist, the broadcast is sent along all paths except the originating one. The other endpoint issues a similar broadcast, and the frame is continuously rebroadcasted, causing a switching loop. Since Layer 2 frames lack a TTL (Time to Live), they can remain in the network indefinitely, and often the resolution involves powering down the affected switches.
Nokia, as well as the broader industry, has implemented two main solutions to address this issue:
This protocol detects and prevents loops in Layer 2 traffic. It accomplishes this by designating one of the participating switches as the root bridge, or root switch, which acts as the master. The other switches assume the roles of backups and subordinates. It also sends some ports into an inactive state to block unnecessary paths, thereby preventing loops.
The root bridge can be selected by the participating bridges or manually configured by the Network Administrator. The selection process is as follows:
There is a severe attack that exploits STP, or more accurately, poorly configured STP. This attack abuses the root selection process by assuming a lower Bridge ID number and taking over as the root bridge, thus becoming the control center for Layer 2 traffic. To protect against this, it is crucial to configure BPDU guard when using STP. It is also advisable to manually select your root bridge rather than leaving it to the switches by default. Additionally, a DDoS (Distributed Denial of Service) attack can be initiated when a rogue switch floods the network with BPDUs.
STP selects the path with the lower cost and blocks the ones with higher costs. The lower bandwidths are the costliest. Cost assignment is typically automatic. Below is the cost mapping:
10g 2 1g 4 622mb 6 155mb 14 100mb 19 45mb 39 16mb 62 10mb 100 4mb 250
In Spanning Tree Protocol (STP) selection or calculations, the following rules are considered:
- The lowest bridge ID is always preferred.
- All ports on the root bridge are designated ports.
- Often, all ports opposite the root bridge's ports are designated ports.
- Depending on the network architecture, at least one or more ports may enter a blocking state.
- Each network segment has one designated port.
- Root ports cannot be designated ports.
- Each participating switch has one root port.
The tie-breaking order is as follows:
- The lowest bridge ID prevails.
- The lowest root path cost takes precedence.
- The lowest MAC address is decisive.
- The lowest port ID is the final determinant.
We have already discussed the root bridge and path selection.
Although not considered as part of the port states, I included this information to review the states that each spanning tree port undergoes, starting just before a new switch connection is established. At this stage, no switch has been connected. No BBDU messages has been exchanged.
Every spanning tree port begins in this state, receiving and processing BPDUs. During the election process, if it becomes the elected root or designated port, it transitions from this state to the Listening state. If not, it remains in the Blocking state.
All root and designated ports transition from the blocking state to this state. Non-designated and non-root ports remain in the blocking state. They receive BPDUs and send them to the switch for processing.
This state is attained after the Listening states. Mac addresses are learnt and used in building a Mac address table. Forwarding is not done at this state
The system forwards frames at a certain state, after which data transmission occurs.
Networking Niche
Copyright © 2024 Networking Niche - All Rights Reserved.
Powered by GoDaddy