"Copyright notice: Original translation articles, translation level is limited, errors are unavoidable, The Translator is not responsible for the consequences of errors or omissions in the article, please be careful to reprint them. Reproduced please retain this statement and source: blog.csdn.net/shallnet, download the English version of the book "
Spanning tree is widely used in the production of a loop-free network topology using two concepts:
- Bridge ID (BID)
- Path overhead
The Bridge ID Bridge ID (BID) is a single 8-byte domain consisting of the two subdomains shown in Figure 6-5.Figure 6-5. Bridge ID (BID) consists of a bridge priority and a MAC address
The low-byte domain consists of a switch 6-byte MAC address,Catalyst 6000 uses a MAC address from the 1024 address pools assigned to the Super Administrator module or backplane, which is not allowed to be modified by the user. The MAC address in the bridge ID is in 16 binary format. Tips:some cataysts from the management module and obtain MAC addresses (such as Catayst 5000), while others are obtained from backplane (such as Catayst 5500 and 6000). bid High-byte domain is the bridge priority, do not confuse the bridge priority and the seventh chapter "Advanced Spanning Tree." The port priority discussed, the bridge priority is a two-byte (16-bit) value, an unsigned 16 is an integer with 216 possible indications, and the range is 0-65535. The default bridge priority is the median value of 32768, which is described using the 10 binary. Note:This book covers only the version of the IEEE Spanning Tree Protocol, although the basic mechanism is the same, there are some differences between IEEE STP and Dec STP (the oldest spanning tree protocol implementation). For example, the SEC STP Bridge priority is 8-bit, and the 2-layer switch catalysts (like 4000s,5000s and 6000s) supports only IEEE STP, which is supported in both versions of Cisco routers.
Path OverheadThe bridge uses the concept of overhead to evaluate its distance from other switches, and 802.1D initially defines the cost of 1000Mbps link bandwidth in Mbps, such as the overhead of a 10BaseT link of 100 (1000/10), The overhead of Fast Ethernet and FDDI is 10 (1000/100). Thismethod of computing has been used until the 1983 when Radia Perlman started the protocol, with the rise of Gigabit Ethernet and OC-48 ATM (2.4 Gbps), the problem of the path overhead cannot be expressed using decimals, such as oc_48 The cost of the ATM is calculated as 1000mbps/2400mbps=0.41667, but this is an invalid path cost value. One solution to this problem is that a link greater than 1Gbps uses 1 as the path cost, but this does not conform to the criteria that STP will always accurately select "Optimal path" in gigabit networks. to solve this problem, the IEEE decided to use a non-linear calculation method to modify the path cost, and table 6-1 lists the new path cost values. Table 6-1. STP cost Values for Network Bridges
bandwidth |
STP cost |
4 Mbps |
250 |
10 Mbps |
100 |
16 Mbps |
62 |
45 Mbps |
39 |
100 Mbps |
19 |
155 Mbps |
14 |
622 Mbps |
6 |
1 Gbps |
4 |
Ten Gbps |
2 |
The values in table 6-1 are well thought out, so that the current and previous methodsare applicable to the bandwidth that is now commonly used. The key point about STP path cost values is to remember that "the smaller the overhead is the better". Remember 1. The X version to the 2.4 version of the Catalyst NMP uses the previous linear value calculation method, starting with the 3.1 release with the new cost value, all Catalyst 4000s and 6000s using the new cost value.
The classic "Cisco Lan Switching" chapter sixth (iii): Double Key spanning-tree Protocol Concepts
