10 Dynamic Route Agreement
10.1 Introduction In the previous chapters, we discussed static elevation. When configuring an interface, the routing table is generated by default, and the entry is added to the Route command (usually from the system self-boot program file), or via the ICMP to change the route generation entry (usually The default method is wrong). In the network, only a single connection point with other networks and there is no excess route (if the main route fails, the spare route can be used), which is feasible. If the above three cases cannot be met, a dynamic selection is usually used. This chapter discusses the dynamic selection protocol, which is used for communication between routes. We mainly discuss RIP, ie the circuit information protocol (Routing Infromation Protocol), and most TCP / IP implementations provide a wide range of protocols. Then we discuss two new selection protocols, OSPF and BGP. In this chapter, a new selection technique named non-classification inter-area selection is studied. Now the INTERNET is starting to adopt the protocol to maintain the number of Class B networks.
10.2 Dynamic Route When communication is performed between adjacent routers to inform the other party's currently connected network, there is a dynamic selection. The router must be used to communicate with the selection protocol, which has many kinds of such selection protocols. There is a process on the router called routing daemon, which runs the routing protocol and communicates with some routers adjacent. As shown in Figure 9.1, the routing daemon updates the routing table in the kernel based on the information received from the adjacent router. Dynamic selection does not change the way we are selected in the IP layer in the IP layer. Our circuit is called the routing mechanism. The kernel search routing table, the way the host route, network routing, and the default route have not changed. It is only the information placed in the routing table changes - when the route changes over time, the route is dynamically increased or deleted by the routing daemon, not from the Route command in the self-boot program file. As we described earlier, the Routing Guards Add Routeing Policy to the system, select routing and add to the routing table. If the daemon discovers that there are multiple routes to the same SBO, it (in some way) will select the best route and join the kernel routing table. If the route daemon finds a link has been disconnected (it may be a router crash or phone line is not good), it can delete the affected route or add another route to bypass the problem. In systems like Internet, there are currently many different selection protocols. Internet is organized in a group of autonomous systems as (Autonomous System), each autonomous system is usually managed by a single entity. A company or university campus is often defined as a self-government system. The NSFNET backbone network forms a self-government system because all routers in the backbone network are under a single management control. Each autonomous system can select the selection protocol between the various routers in the autonomous system. This protocol we call it the internal gateway protocol IGP (Interior Gateway Protocol) or an IntrAdomain Routing Protocol. The most commonly used IGP is the selection information protocol RIP. A new IGP is an open shortest path priority OSPF (Open Shortest Path First) protocol. It is intended to replace RIP. Another IgP protocol used in 1986 in the original NSFNET backbone online - Hello is now no longer available.
(Below is the original book P.1281) The new RFC [Almquist 1993] provides that the router that implements any dynamic selection protocol must support OSPF and RIP, and other IGP protocols can also be supported. Separation of Separated Route Protocols for External Gateway Protocols or Domain Clear Protocols For routers between different autonomous systems. In history, the improved EGP has a protocol that is the same as it names: EGP. The new EGP is currently used on the NSFNET backbone network and some of the regional networks connected to the backbone network that border gateway protocol BGP (Border Gateway Protocol). BGP is intended to replace EGP.
10.3 UNIX Routing Guarding UNIX systems often run named ROUTED routing daemon. This process is provided almost all TCP / IP implementations. This program only uses RIP to communicate, we will discuss this protocol in the next section. This is a protocol used in small to medium networks. Another program is GATED. IGP and EGP support it. [Fedor 1998] describes the early development of Gated. Figure 10.1 compares the different selection protocols supported by ROUTED and two different versions. Most of the system running a route daemon can run routed unless they need other protocols supported by Gated.
Figure 10.1 Routed and Gated
We describe the RIP version 1 in the next section, which describes the difference between the different points, 10.6 of the RIP version 2, which describes OSPF in 10.5, which describes the BGP in Section 10.7.
10.4 RIP: Cruising Information Protocol This section describes RIP because it is the most widely used (also the most attack) selection protocol. The official description file for RIP is RFC 1058 [Hedrick 1988a], but the RFC is only a few years after the protocol is implemented.
The packet format RIP message is included in the UDP datagram, as shown in Figure 10.2. (We described more detailed descriptions in Chapter 11.)
Figure 10.2 RIP packet packaged in UDP datagram
Figure 10.3 shows the RIP packet format when using the IP address. The command field is 1 representation request, and 2 means answered. There are two orders that are discarded (3 and 4), two informal commands: polling (5) and polling entries (6). The request indicates that other systems are required to send all or part of the routing table. Answer, contains all or part of the route table. The version field is usually 1, while the second edition RIP (10.5) sets this field to 2. Keep up with the 20-byte specified address series (for IP addresses), the IP address, and the corresponding metric for the IP address. It can be seen from this section that RIP's metrics are hopped. This 20-byte format of RIP packets can be advertised with up to 25 routes. The upper limit 25 is used to ensure the total length of the RIP packet is 20 × 25 4 = 504, less than 512 bytes. Since each packet carries up to 25 routing, multiple packets often need to send the entire routing table.
Running normally let us look at the results of the ROUTED program that uses the RIP protocol. The UDP port number commonly used by RIP is 520. · Initialization: When starting a routing daemon, it first judges which interfaces are started, and send a request message on each interface, requiring other routers to send a full routing table. In point-to-point link, the request is sent to other endpoints. If the network supports broadcast, this request is sent in broadcast form. The destination UDP port number is 520 (this is the route guardial port number of other routers). This command field of this request message is 1, but the address series field is set to 0, and the metric field is set to 16. This is a special request message that requires the other end routing table. · Receive request. If this request is the special request we just mentioned, the router sends a complete routing table to the requester. Otherwise, each of the entries in the request: If we have routes that are connected to the indicated address, set the metric to our value, otherwise the degree is set to 16. (Metrics 16 is a special value called "infinity", which means that we did not reach the route of the purpose.) And send back response. · Receive response. The routing table may be updated to the response. A new entry may be added to modify existing entries, or delete the already entries. · Regular selection update. For every 30 seconds, all or part of the router sends its full routing table to the neighboring router. The transmission routing table can be in the form of a broadcast (such as in Ethernet), or other endpoints transmitted to the point-to-point link. · Trigger update. Whenever a routing metrics change, it is updated. You don't need to send a full routing table, but only you need to send a changing entry. Each route has a timer associated with it. If the system running the RIP discovers that a route is not updated within 3 minutes, the measurement of the route is set to infinity (16), and is marked as delete. This means that we have not received the update of the router that advertises the router in 6 30 seconds. After another 60 seconds, the route will be removed from the local routing table to ensure that the failure of the route has been propagated. The metric used by metric RIP is calculated in hop (HOP). All direct connection interfaces are 1. Consider the routers and networks shown in Figure 10.4. The four dashed lines we draw are broadcast RIP packets.
Figure 10.4 Router and web example
The router R1 is advertised to N1 by sending broadcast to N1 that the number of hops between N2 is 1. (The route between the N1 is advertised to N1 is useless.) It also advertises the number of hops between N1 while transmitting broadcast to N2. Similarly, R2 advertises its metric from N2 to 1, and the metric with N3 is 1. If the neighboring router is not in line with other network routes, then our metric is 2, because in order to send a message to the network, we must pass the router. In our example, the metric of R2 to N1 is 2, as in the metric of R1 to N3. Since each router sends its routing table to the neighboring station, it can be judged that the same autonomous system AS is routed to each network. If there is a plurality of routes from a router to a network within the AS, the router will select the smallest jump, and ignore other routes. The maximum number of hops is 15, which means that RIP can only be used within AS of the maximum number of hops between the host. The measurement is 16 indicates that the IP address is not routed.
Problem This method looks simple, but it has some defects. First, the RIP does not have the concept of subnet address. For example, if the host number of 16 BIT is not 0 in the standard B address, then RIP cannot distinguish non-zero is a subnet number, or a host address. There are some implementations to use the network mask of the interface through the received RIP information, which is likely to be wrong. Second, after the router or link is faulty, it takes a long period of time to stabilize. This time is usually taken for a few minutes. Routing loops can occur during this setup time. When implementing RIP, there must be a lot of subtle measures to prevent the emergence of routing loops and make it established as soon as possible. RFC 1058 [Hedrick 1988a] pointed out a lot of details that implement RIP. The use of hops is used as a routing metric to ignore other factors that should be considered. At the same time, the maximum metric is 15 limits the size of the network that can use the RIP. Example We will use the RipQuery program to query the routing table in the router, which can be obtained from the Gated. The RipQuery program gives a complete routing table by sending an informal request ("poll" in Figure 10.3 is 5). If the response is not received within 5 seconds, the standard RIP request (1) is sent (1). (We mentioned earlier, set the address line field to 0, measure the number of items to 16, require other routers to send its full routing table.) Figure 10.5 shows that we will query their routing table from the Sun host. Two routers. If we execute the Ripquery program on the host Sun to get the next router Netb's selection information, then we can get the following results:
Sun% ripquery -n netb504 bytes from netb (140.252.1.183): The first message contains 504 bytes -------------------- here deleted many lines 140.252 .1.0, Metric 1 Figure 10.5 above Ethernet 140.252.13.0, Metric 1 Figure 10.5 below Ethernet 244 bytes from netb (140.252.1.183): The second message contains the remaining 244 bytes below to delete many lines
As we guess, Netb tells us that the measurement of the subnet is 1. In addition, Metric (140.252.1.0) of the Ethernet (140.252.1.0) connected to Netb is also 1. (-N parameter represents direct printing IP addresses without need to view its domain name.) In this case, the NetB is configured to think that the host located directly at 140.252.13 subnets is directly connected to it - ie Netb doesn't know which The host is truly connected to 140.252.13 subnets. Since there is only one connection point with the 140.252.13 subnet, the metric of each host is not greatly significant.
Figure 10.5 We will query two routers Netb and Gateway from the routing table content
Figure 10.6 shows a message using TCPDUMP. We use the -i S10 parameter to specify the SLIP interface.
Figure 10.6 TCPDUMP output results running the Ripquery program
The first request issues a RIP polling command (line 1). This request is timeout after 5 seconds, issues a regular RIP request (line 2). Last lines and 2nd rows The last 24 represents the length of the request message: 4 bytes of RIP headers (including commands and versions), and then a single 20-byte address and metric. The third line is the first answer message. The last 25 of the line indicates that 25 addresses and metrics are included, and we have calculated in front, the number of bytes is 504. This is the result of the above RipQuery program. We specify the -s600 option for the TCPDump program to allow it to read 600 bytes from the network. In this way, it can receive the entire UDP datagram (rather than the front half of the packet), then print the contents of the RIP response. We will omit the output result. The fourth line is the second response message from the router, which contains 12 addresses and metric pairs behind. We can calculate that the length is 12 × 20 4 = 244, which is the result of the RipQuery program before. If we cross the Netb router, go to Gateway, you can predict that our subnet (140.252.13.0) is 2. We can run the following command to verify: sun% ripquery -n gateway 504 bytes from Gateway (140.252.1.4): -------------------- here deleted a lot Row ---- 140.252.1.0, Metric 1 Figure 10.5 above Ethernet 140.252.13.0, Metric 1 Figure 10.5 below Ethernet
Here, the measurement of the Ethernet (140.252.1.0) on Figure 10.5 is still 1 because the Ethernet is directly connected to Gateway and NetB. And our subnet 140.252.13.0 is just like expected, its measurement is 2.
Another example we now look at all the RIP updates of all non-active requests on Ethernet to see information that RIP is regularly sent to its neighbors. Figure 10.7 is a variety of arrangements in the NOAO.EDU network. To simplify, we do not use the router used in this article, while RN represents the router, where n is a subnet number. We represent point-to-point links in dashed lines and give the IP address of these link peers.
Figure 10.7 NOAO.EDU 140.252 Multiple Networks
We run Solaris 2.x snoop programs on the host Solaris, which is similar to TCPDUMP. We can run the program without the conditions that need super user privileges, but it only captures broadcasted packets, multicast packets, and packets sent to the host. Figure 10.8 gives messages captured in 60 seconds. Here, we will represent most official host names in RN.
Figure 10.8 Solaris The RIP Broadcast Packet captured in 60 seconds
The -p flag is captured in non-mixed mode, and -tr prints the corresponding time stamp, and the UDP Port 520 captures UDP datagram that is 520. The first 6 packets from R6, R4, R2, R7, R8, and R3, only one network to each message. To view these packets, we can find that R2 advertisements go to a routing of 140.252.6.0 to 1, R4 advertisement to a route for the number of hops of 140.252.4.0, and so on. However, the Gateway router notned 15 routes. We can view all the contents of the RIP packet by adding the -v parameter when running the Snoop program. (This logo outputs all the contents of all packets: Ethernet first, IP head, UDP head, and RIP packets. We only retain RIP information and remove other information.) Figure 10.9 shows the output results. The route passed through these subnet 140.252.1 is compared to the topology in Figure 10.7. One problem that makes people confused is why Figure 10.8 Output results, R10 advertises its 4 networks and only 3 shown in Figure 10.7. If we look at RIP packets with Snoop, you will get the following Note Routing: RIP: Address MetriCrip: 140.251.0.0 16 (Not Reachable) Rip: 140.252.9.0 1Rip: 140.252.11.0 1
The route to the B network 140.251 is fake, should not be advertised. (It belongs to other agencies instead of noao.edu.)
Figure 10.9 RIP response from GATEWAY
In Fig. 10.8, for the RIP packet sent by R10, the SNOOP output "Broadcast" symbol, which indicates that the destination IP address is limited broadcast address 255.255.255.255 (12.2), not other routers to point to the broadcast address of the subnet ( 140.252.1.255).
10.5 RIP protocol version 2RFC 1388 [Malkin 1993A] is expanded, and it is often referred to as RIP-2. These expansion do not change the protocol itself, but use some of the items labeled "must be 0" in Figure 10.3 to deliver some additional information. If the RIP ignores the fields that must be 0, RIP and RIP-2 can be interoperable. Figure 10.10 Reforms the diagram defined by the RIP-2. For RIP-2, its version field is 2. The Routing Domain is an identifier of a selection daemon, which points out the owner of this datagram. In a UNIX implementation, it can be the process number of the selection daemon. This domain allows managers to run multiple RIP instances on a single router, and each instance is running in one selected road domain. The route tag is to support the external gateway protocol. It carries a self-government number of EGP and BGP. The subnet mask of each entry is applied to the corresponding IP address. The next stop IP address indicates where the message sent to the destination IP address. This field means that the message to the destination address should be sent to the system that transmits the RIP packet.
Figure 10.10 RIP-2 message format
RIP-2 provides a simple authentication mechanism. You can specify the first 20-byte entry address series of RIP packets to 0xFFF, and the routing is 2. The remaining 16 bytes in the entry contain a clear text. Finally, in addition to the broadcast (Chapter 12), RIP-2 also supports multicast. This reduces the load of the host that does not listen to the RIP-2 packet.
10.6 OSPF: Open Maximum Path Priority OSPF is another internal gateway protocol except RIP. It overcomes all limitations of RIP. The second edition OSPF is described in RFC 1247 [MOY 1991]. Unlike the RIP protocol used by the distance vector, OSPF is a link status protocol. The distance vector means that the packet sent by the RIP contains a distance vector (number of hops). Each router updates its own routing table based on these distance versions of the neighbors. In a link status protocol, the router does not exchange distance information with its neighbors. It uses that each router actively tests the state of connecting the link connected to its neighboring station, and sends this information to other ingots, and the neighboring information spreads out in the autonomous system. Each router receives these link status information and establishes a complete routing table. From a practical perspective, the difference between the two is that the link status protocol is always faster than the distance vector protocol. Convergence means that after the route changes, for example, after the router is turned off or the link is faulty, it can be stabilized. [Perlman 1992] compared to other aspects of these two types of selection protocols. The difference between OSPF and RIP (and other circuit protocols) is that OSPF uses IP directly. That is, it does not use UDP or TCP. For the Protocol field of the IP head, OSPF has its own value (Figure 3.1). In addition, as a link state protocol instead of a distance vector protocol, OSPF has some features superior to RIP. 1. OSPF can calculate the respective routing sets for each IP service type (Figure 3.2). This means that for any purpose, there are multiple routing tables, each entry corresponding to an IP service type. 2. Assign a widthless fee to each interface. The assignment can be performed by throughput, round trip time, reliability or other performance. A separate fee can be assigned to each IP service type. 3. When there is a plurality of routes of the same amount of time, OSPF allocates traffic on these routes. We call it a flow balance. 4. OSPF Support Subnet: Subnet Mask and Each Notification Routing Connection. This allows one of the types of IP addresses to multiple different sizes. (We give such an example in Section 3.7, called the length subnet.) The route to a host is notified by a total 1 subnet mask. The default route is 0.0.0.0 with an IP address, and the network mask is notified. 5. The point-to-point link between the router does not require an IP address per side. We call them no network. This saves IP addresses - a very shortcoming resource now. 6. A simple authentication mechanism is adopted. A express texttening will be specified using methods similar to the RIP-2 mechanism (10.5). 7. OSPF uses multicast (Chapter 12), not a broadcast form to reduce the system load that does not participate in OSPF. With most vendors support OSPF, OSPF in many networks will gradually replace RIP.
10.7 BGP: Boundary Gateway Protocol BGP is an external gateway protocol for communication between different autonomous systems. BGP is an old EGP resequent product used by ARPANET. RFC1267 [Lougheed and Rekhter 1991] describes the BGP of the third edition. RFC 1268 [Rekhter and GROSS 1991] describes how to use BGP in the Internet. Below, most of the descriptions of BGP are from these two RFC documents. At the same time, the 4th edition of BGP was developed in 1993 (see RFC 1467 [TopolCIC 1993] to support the CIDR described in Section 10.8. The BGP system switches the network between the other BGP systems. This information includes all paths in the autonomous system AS that must be passed through these networks. This information is sufficient to construct a self-government connection diagram. Then, the selection circuit can be deleted according to the connection diagram, and the selection strategy can be set. First, we divide the IP data in a self-government system into this traffic and through traffic. In the autonomous system, local traffic is the flow of starting or terminating from the autonomous system. That is, the host IP address or the host IP address indicated by the IP address is located in the autonomous system. Other flow rates are called through flow. One purpose using BGP in the Internet is to reduce traffic. The autonomous system can be divided into the following types: 1. Subject system (STUB AS), which is only a single connection with other autonomous systems. STUB AS is only local traffic. 2. MultiHomed AS, which has multiple connections to other autonomous systems, but refused to transfer through traffic. 3. Transmission Autonomous System (Transit As) has multiple connections with other autonomous systems, under some policy guidelines, which can transmit local traffic and through traffic. In this way, the Internet's total topology structure can be regarded as any interconnection of some residual autonomous systems, multi-interface autonomous systems, and transfer autonomous systems. The residual pile autonomous system and multi-interface autonomous system do not need to use BGP - they can reach information between autonomous systems by running EGP. BGP allows the use of strategy-based selection. Develop a strategy by the autonomous system administrator and specify the policy to BGP through the configuration file. The development strategy is not part of the protocol, but the specified policy allows the BGP to select the path when there is a plurality of optional paths, and control the resend of information. Cruptory policies are related to political, security or economic factors. The BGP and RIP and OSPF are between BGP use TCP as its transport layer protocol. Two TCP connections between the two systems running BGP and then exchange the entire BGP routing table. From this time, the update signal is sent again when the routing table changes. The BGP is a distance vector protocol, but is different from the RIP (notified to the destination address hop), the BGP lists the route to each destination address (the autonomous system to the serial number of the destination address). This eliminates some of the problems of the distance vector protocol. The 16 bit digit is used to represent the autonomous system identification. BGP Detects the TCP Connection to the TC link or host failure by regularly sending Keepalive packets to its neighbors. The time interval between the two messages is 30 seconds. The Keepalive Packets of the Application Layer with the KeePalive Options for TCP (Chapter 23) are independent.
10.8 CIDR: ClassSs Interdomain Routing In Chapter 3, we pointed out the lack of the Class B address, so many network sites can only use multiple Class C network numbers without using a single Class B network number. Although allocate these Class C addresses solve a problem (the lack of the class B address), it brings another question: Each Class C network requires a routing table entry. There is no type of domain selection (CIDR) is a method of preventing Internet routing tables. It also refers to Supernetting, which is described in RFC 1518 [Rekher and Li 1993] and RFC 1519 [Fuller et al. 1993], [ford, rekhter, and Braun 1993] is a review . CIDR has an InternetAtAtitecture Board's Blessing [huitema 1993]. RFC 1467 [TopolCIC 1993] summarizes the CIDR development status in the Internet. The basic view of CIDR is to adopt a way to allocate multiple IP addresses, making them more in summarization in the routing table. For example, if 16 Class C addresses can be assigned to a single site, one of these 16 addresses can be assigned in a manner that can be used, so that all 16 addresses can refer to the single routing table entry on the Internet. At the same time, if there are 8 different sites to access the Internet with the same connection point of the same Internet service provider, and the 8 different IP addresses allocated by this 8 sites can be summarized, then for these 8 sites. On the Internet, only a single routing table item is required. To use this sum, you must meet the following three characteristics. 1. These IP addresses must have the same high address bits for the sum of the plurality of IP addresses to be selected for multiple IP addresses. 2. The routing table and the selection algorithm must be extended to make a selection decision based on the 32 Bitip address and the 32 BIT mask. 3. You must extend the selection protocol to have a 32 bit mask in addition to the 32 bit address. OSPF (10.6) and RIP-2 (10.5) are able to carry the 32 BIT mask raised by the 4th edition BGP. For example, RFC 1466 [Gerich 1993] It is recommended that the new Class C address in Europe is 194.0.0.0 to 195.255.255.255. Taking 16 binders, these addresses range from 0xc2000000 to 0xc3FFFFF. It represents 65,536 different Class C network numbers, but their address is 7 bits is the same. In countries other than Europe, IP address can be used to selection of all of these 65536 Class C network numbers to a single point for all of these 65536 Class C network numbers. The rear of the Class C address (i.e., each of the bits behind 195 or 195) can also be assigned, such as the distribution of national or service providers to allow the use of these 32 bit masks between the European routers. 7 other bits outside Bit are summarized. CIDR also uses a technique to make the best match always the longest match: that is, in the 32 bit mask, it has the maximum value. We continue to adopt the examples used in the previous paragraph, and a service provider in Europe may adopt a different access point different from other European service providers.
If the address group assigned to the provider is from 194.0.16.0 to 194.0.31.255 (16 C-network numbers), then only the IP address of the routing entry for these networks is 194.0.16.0, the mask is 255.255.240.0 (0xfffff000). Detailed datagram to 194.0.22.1 will match this routing table entry to the table item of other European C class addresses. However, since the mask 255.255.240 is "long" than 254.0.0.0, a routing table item with a longer mask will be employed. "None type" means that the current circuit decision is based on a mask operation of the entire 32 Bitip address. Regardless of its IP address, there is no difference between Class B or Class C. CIDR is initially proposed for the new Class C address. This change will slow the speed of the Internet routing table, but there is no help for existing elevations. This is a short-term solution. As a long-term solution, if the CIDR is applied to all IP addresses, and the existing IP address is reassigned according to the border and service providers (and all existing hosts are re-addressed!), Then [Ford, Rekhter, And Braun 1993] is claimed that the routing table currently contains 10,000 network entries will be reduced to only 200 entries. 10.9 Summary There are two basic circuit protocols, namely the internal gateway protocol (IGP) between the same autonomous system, and an external gateway protocol (EGP) for router communication within different autonomous systems. The most commonly used IGP is the Routing Information Protocol (RIP), and OSPF is a new IGP that is being widely used. A newly colored EGP is a Boundary Gateway Protocol (BGP). In this chapter, we consider the type of packets in RIP and their exchange. The second edition is RIP is its nearest improvement version, which supports subnets, and some other improvement technologies. I also flashed the OSPF, BGP, and Non-Type Domain Selection (CIDR), CIDR is a new technology, which uses it to reduce the size of the Internet routing table. You may also encounter some other OSI selection protocols. The Domain Room Cruptory Protocol (IdRP) is the first BGP version in order to use an OSI address instead of an IP address. The Intermediate System TO Intermediate System Protocol (IS-IS) is the standard IGP of OSI. It can be used to route CLNP (no connection network protocol), which is an OSI protocol similar to IP. IS-IS and OSPF are similar. Dynamic routine is still a research hotspot in a network interconnection. It is a complex job for the selection protocol and the running route daemon. [Perlman 1992] provides many details.
Exercise: 10.1 What routes in Figure 10.9 enters Gateway from the router KPNO? 10.2 Suppose a router uses the RIP notification 30 routing, which requires a data report that contains 25 routing and another contains 5 routing. If every hour, the first datagram contains 25 routing is lost once, then what is the result? 10.3 There is a checksum field in the OSPF packet format, and the RIP packet does not have this, why? 10.4 Load balancing like OSPF, what is the impact of the transport layer? 10.5 Review RFC 1058 About other information on implementing RIP. In Figure 10.8, each router of the 140.252.1 network is only informing the route it provides, and it does not know any other route through other routers. What is the name of this technology? 10.6 In Section 3.4, we said that there are more than 100 hosts on the 8 routers shown in Figure 10.7, and there are more than 100 hosts on the 140.252.1 subnet. So how do these 100 hosts handle 8 broadcast information every 30 seconds (Figure 10.8)? 10-1
