Friday, March 3, 2017

Leave a Comment

Review of Route Redistribution

  
                            We learned about the concept of Route Redistribution back in the Route Complete Course and we said that Route Redistribution allows routes that learned from one Routing Source
Ø  Maybe like Routing Protocol
Ø  Maybe a directly connected Routes
Ø  Maybe a Statically configured Route
But we learned Routes from some source and those Routes can be injected into a specific Routing Protocol for example, we could Redistribution Routes from OSPF into EIGRP and before we get into Troubleshooting Route Redistribution, let’s take a few movements and review the basics.
First why would we want to do Route Redistribution?
Ø  Well one reason we might want to do Route Redistribution is we got an enterprise network and its running more than one Interior Gateway Protocol
Ø  Maybe because there is a company merger, Company A merge with Company B and one was running OSPF and other was running RIPv2, we might need to Redistribute between those 2 different Interior Gateway Protocols
Ø  Maybe we got different departments within the enterprise that are under different Administrative Control or perhaps we got a Partner Company, it’s not our Enterprise but it’s a Company we do a lot of work with maybe we want to connect out to those partner networks
Ø  We don’t have to do Route Redistribution just between Interior Gateway Protocols, we could Redistribute between an Interior Gateway Protocol and an Exterior Gateway Protocol specifically BGP Border Gateway Protocol
Ø  Maybe we want to advertise our Routes out of Server Provider and we got a BGP Peering Relationship with that Service Provider
Ø  Maybe the reverse we want to take BGP learned Routes and Redistribute those or maybe a subset of those into our IGP
                   As an example, let’s think about this Router it’s got one interface in one Autonomous System and another interface in that different Autonomous System the interface is on the left is connected into EIGRP Autonomous System and interface on the right side connected into OSPF Autonomous System
                              What we can do is take Routes known to EIGRP and inject those into OSPF, similarly we can take Routes known to OSPF and inject those into EIGRP that would be an example of “Mutual Route Redistribution”
                        That’s where we got a couple of different Routing Protocols and each Routing Protocol is sending its Routes into the other Routing Protocol and when we configure Route Redistribution, here is key point to keep in mind often time it can get confusing in which direction the Redistribution is happening.
                      If i go into Router EIGRP 100 Configuration Mode, i am in the Router Configuration for an EIGRP Autonomous System and i give the Redistribute command under EIGRP Router Configuration Mode what i am saying, am i saying that i want to take EIGRP learned Routes and Redistribute them somewhere else.
                     No! What i am saying with Redistribute command is “Redistribute Routes from the specified Routing Source their we gonna specifying in Redistribution command, redistribute those into this Routing Protocol” in my case, i am in EIGRP Router Configuration Mode.
                   I say Redistribute, maybe i am redistributing OSPF, i am Redistributing those OSPF Routes into EIGRP, when i say that i want to Redistribute routes learned by a specific Routing Protocols it’s not just Routes learned by that Protocol, it also includes Routes to networks on directly connected interfaces and those interfaces are enabled for this Routing Protocol those will be Redistributed as well
                  One of the big challenges we have with Redistribution is Routing Protocol use different metrics they got different criteria for what makes up a metric
Ø  OSPF’s metric is based on Cost(Bandwidth).
Ø  And by default, EIGRP’s metric based on Bandwidth and Delay, even though it can consider other things depending on our K Values.
Ø  RIP its metric is based on Hop Count, how many Routers do we have to transit in order to get to this network.
And it really doesn’t make any sense to try to do something like this, we could not say for example, if an OSPF Routes has a metric of 10 and we were Redistributing it into RIPv2 then give it a Hop Count of 3, it really doesn’t make sense to do something like that because these Routing Protocols are calculating their metrics very very differently so, what we can do. There are some default Seed metric
                   “A Seed Metric can be assigned by default to our Route as its being Redistributed into a Routing Protocol” that’s if we don’t manually configure a metric ourselves which we can do, and different Routing Protocol have different Seed Metric and i want you to understand what’s those are, lets populates the Table to see what default Seed Metric are for various Routing Protocols
Ø  The default Seed Metric for RIP is Infinity, if we Redistribute a Route into RIP by default RIP is gonna consider that Route Unreachable
Ø  The default Seed Metric for EIGRP also Infinity
Ø  However, its’ bit different with OSPF, with OSPF if we Redistribute the Route into OSFP that Route is going to inherit a default Seed Metric of 20 unless we Redistributing from BGP that’s an exception if the Route is being Redistributed from BGP, in that case it’s gonna have an OSPF metric of 1
Ø  And BGP is an interesting one, its gonna use whatever the IGP Metric value is as the default Seed Metric.
Now the we reviewed some of the fundamentals of Route Redistribution
Ø  We discussed why we need it?
Ø  What does it do?
Ø  What are the defaults Seed Metric?
                     Let’s go to interface and review a Basic Route Redistribution Configuration and then in our next session we get into Troubleshooting Route Redistribution. Here in Router R2
       We gonna configure the Route Redistribution because it’s the boundary Router it’s got an interface in the OSPF Autonomous System and it’s got another interface in EIGRP Autonomous System this is the place we gonna set things up, let’s see what the Router Configuration looks like right now in R2
Ø  R2#show run | s router
             And we see that for EIGRP interface serial 1/1, if we look at the topology is participating in EIGRP and the OSPF Autonomous System includes our loopback 0 interface as well as serial 1/0. Now let’s set up Mutual Router Redistribution let’s start by Redistributing OSPF learned Router into EIGRP
Ø  R2(config)#router eigrp 1                 
Ø  R2(config-router)#redistribute ospf 1
Now i am gonna setup default metric which will apply to all routes being redistributed into EIGRP
Ø  R2(config-router)#default-metric 1554 10 255 1 1500
ü  1554: Bandwidth (Kbps)
ü  10: Delay (Microseconds)
ü  255: Reliability
ü  1: Load
ü  1500: MTU (Bytes)
                                               Remember the metric for EIGRP is made up of “bandwidth, delay Reliability, Load and MTU (where MTU is not really used in calculation)” but its designed to be Tie-Breaker and by default only Bandwidth and Delay is considered but we can manipulate the K Values to make “Load and Reliability” also be considered here.
                      Once we done that, we should now be Redistributing OSPF known Routes into EIGRP, let’s see if that True, let’s go to R3 and see if we learned anything
Ø  R3#show ip route
                       We have learned about the Loopback IP Address on R1, the Loopback IP Address on R2 we have learned about the network connected of fastetherent 0/0 on R1 and we have learned about network interconnecting R1 and R2, Great News! However, just because R3 knows about these networks how to get to those networks, that’s no guarantee that we can setup conversation with a host on one of those networks for example, let’s try a ping
Ø  R3#ping 1.1.1.1
                     It’s in the IP Routing Table but the ping is gonna failed what do they fail? i am getting to 1.1.1.1 however, that Router R1 it does not know how get back to me and as result all of the Pings failed. We have only Redistributed in one direction, we have Redistributed OSPF into EIGRP let’s go to other direction now, let’s go to R2
Ø  R2(config)#router ospf 1
Ø  R2(config-router)# Redistribute eigrp 1
                      I could say that i wanted to specify a metric in this point but remember OSPF does have a default metric it going to assign to Redistributed Routes, it’s gonna be a metric of 20 unless that Redistributed Routes is coming from BGP in that case metric is gonna be 1 but here i am just gonna leave it at default of 20.
                         Something that we typically want to do is include the “Subnets” keyword, the “Subnets” keyword cause us to Redistribute all ospf knows Routes into EIGRP, if we do not give the “Subnets” keyword we only gonna be Redistributing classful networks infect it get show us warning that “only Classful network are going to be Redistributed”, let’s prove that
Let’s take a look at our OSPF database
Ø  R2#show ip ospf database
                    I am looking for Type 5 LSA’s, a Type 5 LSA would represent a network that we learned another Autonomous System and I don’t see any right now. Just as experiment, let’s create another network in R3 in the EIGRP Autonomous System that does have a classful subnet mask, i am going to R3 and just as a test let’s create Loopback1
Ø  R3(config)#interface loopback 1
And i am assigning Classful Subnet Mask
Ø  R3(config-if)#ip address 4.4.4.4 255.0.0.0
                  This is just an experiment now see let’s that network that classful network, did it get Redistributed into the OSPF, let’s go back to R2
Ø  R2#show ip ospf database
                 Look at this we got Type 5 LSA representing that Classful network, so this point we will Redistributed in Classful networks but we don’t have any other network that have a classful mask, how do we fix that let’s go back to Router 2
Ø  R2(config)#router ospf 1
Ø  R2(config-router)# Redistribute eigrp 1 subnets
Once i do that let’s take a look at OSPF database
Ø  R2#show ip ospf database
                    Great News! We have plenty of Type 5 LSA’s now we know about all of these networks at this point all three should be able to ping the Loopback interface on Router R1
Ø  R3#ping 1.1.1.1
Let’s go to R1
Ø  R1#show ip route
                          And the E2 Codes tells us these are the networks learned them an external Autonomous System
                         E2 is the default as we saw on R1 but the difference between E1 and E2 is “when a Route is being Redistributed as E2 or Type 2 External Route then it’s metric is gonna remain as the same” it’s gonna be the metric assigned by the ASBR, by the Autonomous System Boundary Router regardless of how many router Hop’s it has to go through within an OSPF Areas it still gonna have that original metric assigned by the ASBR, if i am 3 Router Hop’s away and i say what’s the metric to get back this network, it’s gonna be the same metric as the ASBR reported.
You might think that’s not terribly accurate, i want to consider the cost to get from this router back to the ASBR, If you want to consider the metric to get from a router to the ASBR and add that Cost on to the Cost Advertised by the ASBR that’s where you need to set the metric-type to 1(E1)
Ø  (config-router)#redistribute eigrp 1 subnets metric-type 1
That would be an E1 instead of the E2 that we have by default and E1 route it will consider the distance to get to the ASBR and then add on to that Distance of Cost to get to the Destination network

And now the we reviewed the basic Route Redistribution Configuration, let’s take a look at our next session at Trouble Ticket.


If You Like the Post. Don’t forget 
            to “Subscribe/Share/Comment”. Thank You


0 comments:

Post a Comment