Today we will continue our study of section 2.6 of the ICND2 course and look at configuring and verifying the EIGRP protocol. Setting up EIGRP is very simple. As with any other routing protocol such as RIP or OSPF, you enter the router's global configuration mode and enter the router eigrp <#> command, where # is the AS number.
This number must be the same for all devices, for example, if you have 5 routers and they all use EIGRP, then they must have the same autonomous system number. In OSPF, this is the Process ID, or process number, and in EIGRP, the autonomous system number.
In OSPF, to establish a neighborhood, the Process IDs of different routers may not match. In EIGRP, the AS numbers of all neighbors must match, otherwise the neighborhood will not be established. There are 2 ways to enable the EIGRP protocol - without specifying a reverse mask or with specifying a wildcard mask.
In the first case, the network command specifies a classful IP address like 10.0.0.0. This means that any interface with the first octet of the IP address 10 will participate in EIGRP routing, that is, in this case, all class A addresses of network 10.0.0.0 are involved. Even if you enter an exact subnet like 10.1.1.10 without specifying a reverse mask, the protocol will still convert it to an IP address like 10.0.0.0. Therefore, note that the system will accept the address of the specified subnet anyway, but will consider it a classful address and will work with the entire network of class A, B or C, depending on the value of the first octet of the IP address.
If you want to run EIGRP on the 10.1.12.0/24 subnet, you will need to use a reverse mask command network 10.1.12.0 0.0.0.255. Thus, EIGRP works with classful networks without a reverse mask, and with classless subnets, the use of a wildcard mask is mandatory.
Let's move on to Packet Tracer and use the network topology from the previous video tutorial, on the example of which we got acquainted with the concepts of FD and RD.
Let's set up this network in the program and see how it will work. We have 5 routers R1-R5. Even though Packet Tracer uses routers with GigabitEthernet interfaces, I manually changed the network bandwidth and delays so that this scheme coincided with the topology discussed earlier. Instead of the 10.1.1.0/24 network, I connected a virtual loopback interface to the R5 router, which I assigned the address 10.1.1.1/32.
Let's start by setting up the R1 router. I haven't enabled EIGRP here yet, but just assigned an IP address to the router. With the config t command, I enter global configuration mode and enable the protocol by typing router eigrp <autonomous system number>, which must be between 1 and 65535. I select number 1 and hit enter. Further, as I said, you can use two methods.
I can type network and the IP address of the network. Networks 1/10.1.12.0, 24/10.1.13.0 and 24/10.1.14.0 are connected to router R24. They are all on the "tenth" network, so I can use one generic network 10.0.0.0 command. If I press Enter, EIGRP will be started on all three interfaces. I can verify this by issuing the do show ip eigrp interfaces command. We see that the protocol is running on 2 GigabitEthernet interfaces and one Serial interface, to which the R4 router is connected.
If I run the do show ip eigrp interfaces command again to check, I can verify that EIGRP is indeed running on all ports.
Let's go to router R2 and start the protocol using the config t and router eigrp 1 commands. This time we will not use the command for the entire network, but apply a reverse mask. To do this, I enter the network 10.1.12.0 0.0.0.255 command. To check the configuration, use the do show ip eigrp interfaces command. We can see that EIGRP is running only on the Gig0/0 interface, because only this interface matches the parameters of the command entered.
In this case, the reverse mask means that EIGRP mode will be valid for any network whose first three octets of the IP address are 10.1.12. If a network with the same parameters is connected to some interface, then this interface will be added to the list of ports on which this protocol is running.
Let's add another network with the network 10.1.25.0 0.0.0.255 command and see how the list of interfaces that support EIGRP will now look like. As you can see, now we have added the Gig0/1 interface. Note that the Gig0/0 interface has one peer, or one neighbor, the router R1, which we have already configured. Later, I will show you the commands to check the settings, while we continue to configure EIGRP for the rest of the devices. We may or may not use a backmask when configuring any of the routers.
I go to the CLI console of the R3 router and in the global configuration mode I type the commands router eigrp 1 and network 10.0.0.0, then I go into the settings of the R4 router and type the same commands without using the backmask.
You can see how EIGRP is easier to configure than OSPF - in the latter case, you need to pay attention to ABRs, zones, determine their location, etc. None of this is required here - I just go to the global settings of the R5 router, type router eigrp 1 and network 10.0.0.0, and now EIGRP is running on all 5 devices.
Let's take a look at the information we talked about in the last video. I go into the R2 settings and type the show ip route command and the system shows the required entries.
Let's pay attention to the R5 router, or rather, to the network 10.1.1.0/24. This is the first line in the routing table. The first number in brackets is the administrative distance, which is 90 for the EIGRP protocol. The letter D means that the information about this route is provided by the EIGRP protocol, and the second number in brackets, equal to 26112, is the metric of the route R2-R5. If we go back to the previous diagram, we see that here the metric value is 28416, so I have to see what is the reason for this mismatch.
We type the show interface loopback 0 command in the R5 settings. The reason is that we used a loopback interface: if you look at the R5 delay on the diagram, then it is 10 μs, and in the router settings we are given information that the DLY delay is 5000 microseconds. Let's see if I can change this value. I go into R5 global configuration mode and type interface loopback 0 and delay commands. The system gives a hint that the delay value can be assigned in the range from 1 to 16777215, and in tens of microseconds. Since the delay value of 10 μs corresponds to 1 in tens, I enter the delay 1 command. We check the interface parameters again and see that the system did not accept this value, and it does not want to do this even when updating the network parameters in the R2 settings.
However, I assure you that if we recalculate the metric for the previous scheme, taking into account the physical parameters of the R5 router, the feasible distance value for the route from R2 to the 10.1.1.0/24 network will be 26112. Let's look at the similar values in the R1 router parameters by typing the command show ip route. As you can see, for the 10.1.1.0/24 network, a recalculation was made and now the metric value is 26368, not 28416.
You can check this recalculation based on the scheme from the previous video tutorial, taking into account the peculiarities of Packet Tracer, which uses different physical parameters of the interfaces, in particular, a different delay. Try to create your own network topology with these bandwidth and latency values and calculate its parameters. In your practice, you will not need to perform such calculations, just know how to do it. Because if you want to use the load balancing we mentioned in the last video, you need to know how you can change the delay. I do not advise touching the bandwidth, to adjust EIGRP it is quite enough to change the delay values.
So, you can change the values of bandwidth and delay, thereby changing the values of the EIGRP metric. This will be your homework. As usual, for this you can download from our website and use both network topologies in Packet Tracer. Let's go back to our scheme.
As you can see, configuring EIGRP is very simple, and you can use two ways to designate networks: with or without a backmask. As in OSPF, in EIGRP we have 3 tables: the neighbor table, the topology table, and the route table. Let's take a look at these tables again.
Let's go into the R1 settings and start with the neighbors table by entering the show ip eigrp neighbors command. We see that the router has 3 neighbors.
Address 10.1.12.2 is router R2, 10.1.13.1 is router R3 and 10.1.14.1 is router R4. The table also displays through which interfaces communication with neighbors is carried out. The Hold Uptime is shown below. If you remember, this is the time period, which defaults to 3 Hello periods, or 3 x 5s = 15s. If during this time a Hello response has not been received from the neighbor, the connection is considered lost. Technically, if the neighbors answer, this value goes down to 10s and then back to 15s. Every 5 seconds, the router sends a Hello message, and the neighbors respond to it within the next five seconds. The round-trip time for SRTT packets is given as 40 ms. Its calculation is performed by the RTP protocol, which EIGRP uses to organize communication between neighbors. And now we will look at the topology table, for which we use the show ip eigrp topology command.
The OSPF protocol describes in this case a complex, deep topology that includes all routers and all links available on the network. The EIGRP protocol displays a simplified topology based on two route metrics. The first metric is the minimum feasible distance, which is one of the characteristics of the route. Further, through the slash, the reported distance value is displayed - this is the second metric. For network 10.1.1.0/24, which is connected to router 10.1.12.2, the feasible distance value is 26368 (first value in parentheses). The same value is placed in the routing table because router 10.1.12.2 is the receiver - Successor.
If the reported distance of another router, in this case the value 3072 of router 10.1.14.4, is less than the feasible distance of the nearest neighbor, then that router is a Feasible Successor. If communication with router 10.1.12.2 is lost via the GigabitEthernet 0/0 interface, router 10.1.14.4 will take over the Successor function.
In OSPF, calculating a route through a backup router takes a certain amount of time, which, with a significant network size, plays a significant role. EIGRP does not waste time on such calculations, because it already knows a candidate for the role of Successor. Let's take a look at the topology table using the show ip route command.
As you can see, it is Successor, that is, the router with the lowest FD value, that is placed in the routing table. The channel with metric 26368 is indicated here, which is the FD of the destination router 10.1.12.2.
There are three commands that can be used to check the routing protocol settings for each interface.
The first is show running-config. Using it, I can see what protocol is running on this device, this is indicated by the router eigrp 1 message for network 10.0.0.0. However, from this information it is impossible to determine on which interfaces this protocol is running, so I have to look at the list with the parameters of all R1 interfaces. At the same time, I pay attention to the first octet of the IP address of each interface - if it starts with 10, then EIGRP is in effect on this interface, since in this case the condition for matching with the network address 10.0.0.0 is satisfied. Thus, using the show running-config command, you can find out which protocol is running on each interface.
The next test command is show ip protocols. After entering this command, you can see that the routing protocol is "eigrp 1". Next, the values of the K coefficients for calculating the metric are displayed. Their study is not included in the ICND course, so in the settings we will accept the default K values.
Here, as in OSPF, the Router-ID is displayed as an IP address: 10.1.12.1. If you do not assign this parameter manually, the system automatically selects the loopback interface with the highest IP address as the RID.
The following indicates that automatic route summarization is disabled. This is an important point, because if we are using subnets with classless IP addresses, it is better to disable summation. If you enable this feature, the following will happen.
Imagine that we have routers R1 and R2 using EIGRP, and 2 networks are connected to router R3: 10.1.2.0, 10.1.10.0 and 10.1.25.0. If autosummary is enabled, then when R2 sends an update to R1, it indicates that it is connected to the 10.0.0.0/8 network. This means that all devices connected to the 10.0.0.0/8 network send updates to it, and all traffic destined for the 10. network must be directed to R2.
What happens if another router R1 connected to the networks 3 and 10.1.5.0 is connected to the first router R10.1.75.0? If R3 is also using autosummary, it will tell R1 that all traffic destined for the 10.0.0.0/8 network should be directed to it.
If R1 is connected to R2 on 192.168.1.0 and R3 is connected to 192.168.2.0, then EIGRP will only make autosummary decisions at the R2 layer, which is incorrect. Therefore, if you want to use autosummary for a specific router, in our case R2, make sure that all subnets with the first octet of the IP address 10. are connected only to this router. You must not have networks connected 10. somewhere else, to another router. A network administrator who intends to use autosummarization of routes must ensure that all networks with the same classful address are connected to the same router.
In practice, it is more convenient that the autosum function is disabled by default. In this case, router R2 will send separate updates to router R1 for each of the networks connected to it: one for 10.1.2.0, one for 10.1.10.0, and one for 10.1.25.0. In this case, the R1 routing table will be replenished with not one, but three routes. Of course, summarization helps to reduce the number of entries in the routing table, but if you misplan it, you can destroy the entire network.
Let's go back to the show ip protocols command. Note that here you can see the Administrative Distance value of 90, as well as the Maximum path for load balancing, which defaults to 4. All of these paths have the same cost. Their number can be reduced, for example, to 2, or increased to 16.
Next, the maximum size of the hop counter, or routing segments, is 100, and the value is Maximum metric variance = 1. In EIGRP, the Variance variance allows you to consider equal routes whose metrics are relatively close in value, which allows you to add several routes with unequal metrics to the routing table leading to the same subnet. We'll look at this in more detail later.
The information Routing for Networks: 10.0.0.0 is an indication that we are using the no backmask option. If we go into the R2 settings, where we used the reverse mask, and enter the show ip protocols command, we will see that Routing for Networks for this router is two lines: 10.1.12.0/24 and 10.1.25.0/24, that is, there is an indication of the use of a wildcard mask.
For practical purposes, you don't have to remember what kind of information the test commands give - just use them and view the result. However, on the exam, you will not have the opportunity to answer the question, which can be checked with the show ip protocols command. You will have to choose one correct answer from several options. If you are going to become a high-level Cisco specialist and receive not only a CCNA certificate, but also a CCNP or CCIE, then you need to know what specific information this or that test command produces and what the execution commands are for. You must master not only the technical part of Cisco devices, but also understand the Cisco iOS operating system in order to properly configure these network devices.
Let's return to the information that the system issues in response to the show ip protocols command. We see Routing Information Sources, represented as lines with an IP address and an administrative distance. Unlike OSPF information, EIGRP does not use Router ID in this case, but the IP addresses of routers.
The last command that allows you to directly view the status of the interfaces is show ip eigrp interfaces. If you enter this command, you can see all the router interfaces that are running EIGRP.
Thus, there are 3 ways to make sure that the device is running the EIRGP protocol.
Let's look at equal cost load balancing, or equal load balancing. If 2 interfaces have the same cost, they will be load balanced by default.
Let's use Packet Tracer to see what it looks like using the network topology we already know. Let me remind you that the bandwidth and delay values are the same for all channels between the depicted routers. I enable EIGRP mode for all 4 routers, for which I go into their settings in turn and type the commands config terminal, router eigrp and network 10.0.0.0.
Suppose we need to choose the optimal route R1-R4 to the loopback virtual interface 10.1.1.1, while all four links R1-R2, R2-R4, R1-R3 and R3-R4 have the same cost. If you enter the show ip route command in the R1 CLI, you can see that the 10.1.1.0/24 network can be reached via two routes: through a 10.1.12.2 router connected to the GigabitEthernet0/0 interface, or through a 10.1.13.3 router connected to interface GigabitEthernet0/1, and both of these routes have the same metrics.
If we issue the show ip eigrp topology command, we will see the same information here: 2 Successor receivers with the same FD value of 131072.
So, we have learned what is equivalent ECLB load balancing, which can be performed both in the case of OSPF and in the case of EIGRP.
However, EIGRP also has unequal-cost load balancing (UCLB), or unequal balancing. In some cases, the metrics may differ slightly from each other, which makes the routes almost equivalent, in which case EIGRP allows load balancing through the use of a value called "variation" - Variance.
Imagine that we have one router connected to three others - R1, R2 and R3.
Router R2 has the lowest FD=90, so it acts as a Successor. Consider the RD of the other two channels. R1's RD of 80 is less than R2's FD, so R1 acts as a backup Feasible Successor. Because R3's RD is greater than R1's FD, it can never become a Feasible Successor.
So, we have a router - Successor and a router - Feasible Successor. You can place R1 in the routing table using different variance values. In EIGRP, by default, Variance = 1, so the router R1 as Feasible Successor is not in the routing table. If we use the value Variance =2, then the FD value of router R2 will be multiplied by 2 and will be 180. In this case, the FD of router R1 will be less than the FD of router R2: 120 < 180, so router R1 will be placed in the routing table as Successor 'a.
If we equate Variance = 3, then the FD value of the receiver R2 will be 90 x 3 = 270. In this case, the router R1 will also fall into the routing table, because 120 < 270. Do not be embarrassed that the router R3 does not get into the table despite the fact that its FD = 250 with Variance =3 will be less than the FD of router R2, since 250 < 270. The fact is that for router R3 the condition RD < FD Successor is still not met, since RD= 180 is not less, but more than FD = 90. Thus, since R3 cannot initially be a Feasible Successor, even with a variation value of 3, it still will not get into the routing table.
Thus, by changing the value of Variance, we can use unequal load balancing to include the route we need in the routing table.
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