Comparing OSPF with EIGRP
#encor#350-401#ccnp#ccnp enterprise#ccna#200-301#cisco#cisco cert#ospf#eigrp##kwtrain
55.0K Views｜5 Summarized｜6 year ago
💫 Summary
The video compares OSPF and EIGRP, highlighting their differences in topology mapping, cost calculation, and load balancing, while explaining OSPF's link-state approach and EIGRP's flexibility in metrics and decision-making.
✦
A high-level comparison between OSPF and EIGRP is presented.
00:00The topic is relevant for the Cisco encore exam (358-401).
OSPF is emphasized more in the CCNA blueprint compared to EIGRP.
EIGRP will be covered in more detail during the ANRC training for CCMP Enterprise certification.
✦
Link state routing protocols utilize the Dijkstra algorithm for efficient pathfinding.
03:23Link state protocols maintain a complete map of the network topology.
They mathematically determine the lowest cost paths between nodes.
OSPF and IS-IS are examples of link state protocols using the Dijkstra algorithm.
In contrast, distance vector protocols like RIP only provide next hop information without a complete topology map.
✦
OSPF operates like assembling a puzzle through the sharing of link-state advertisements.
06:48Different routers share information about their parts of the network using LSAs.
The collective information allows all routers to create a unified view of the network topology.
Administrative distance determines the reliability of routing information from various sources.
Lower administrative distance values indicate more believable routing sources.
✦
OSPF calculates routing costs based on link bandwidth rather than just hop count.
10:10OSPF prefers paths with higher bandwidth links, calculating costs based on the reference bandwidth.
For a 100 Meg link, the cost is 1, while a 10 Meg link has a cost of 10.
The preferred path from R1 to R2 to R3 has a total cost of 3, compared to a cost of 11 for R1 to R3 to PC2.
OSPF does not allow decimal values for costs, which can complicate calculations with higher bandwidth links.
✦
Delay, reliability, and load are key metrics influencing EIGRP routing decisions.
13:37Delay is cumulative and can affect overall routing performance.
Reliability is measured on a scale of 0 to 255, with 255 indicating 100% reliability.
Load indicates how congested a link is, also rated over 255, with lower values representing less congestion.
The maximum transmission unit (MTU) can serve as a tie breaker but does not factor into the EIGRP metric formula.
✦
EIGRP calculates the best path based on reported distances and costs to neighbors.
17:01R3 evaluates routes to the network 10.1.1.10 from R1 and R2.
Both R1 and R2 report a metric of 1,000 to reach the network.
R3 adds the cost to reach R1 (10,000) and R2 (5,000) to their reported metrics to determine feasible distances.
The lowest feasible distance (6,000 via R2) is selected as the best path.
✦
OSPF and EIGRP handle neighbor communication and load balancing differently.
20:25OSPF has a dead interval, after which it considers a neighbor down if no updates are received.
EIGRP allows neighbors to wait for a specific time before declaring a neighbor down.
OSPF only balances traffic across equal-cost paths, potentially wasting bandwidth.
EIGRP can use a variance feature to send traffic over multiple paths, even if they have different costs.
00:00[Music]
00:04hey everybody welcome back to the
00:06channel this is Kevin and in today's
00:07video we're going to take a look at a
00:09topic from the encore exam that's a
00:12Cisco exam number 358 - 4 0 1 and
00:15specifically the topic is comparing OSPF
00:19with AI GRP I think it's really
00:21interesting that if you look at the new
00:24CCNA blueprint there is zero mention of
00:27AI GRP and when you get into the encore
00:31course which is what most people will
00:33take as their first exam if they're
00:35going after a CCAP enterprise there is
00:37one mention of eigrp and it's today's
00:40topic it's simply comparing EIGRP with
00:43OSPF but if you go take the an RC exam
00:46and that's what a lot of people will
00:48follow up on core width to finally earn
00:50their ccp enterprise you've got to know
00:53quite a bit about eigrp there so in our
00:55an RC training at you bet we'll get into
00:58the details of how to set it up and all
01:00the different features but for today
01:01let's take a look at this single
01:03instance of EIGRP on the Encore
01:06blueprint and that's comparing OSPF with
01:09EIGRP and if you enjoy the video do the
01:12channel a favor if you would and click
01:14the like button and also subscribe so
01:16you don't miss any of our weekly videos
01:18here's our video on comparing the OSPF
01:21with erp enjoy in this video we want to
01:25compare OSPF with erp both our
01:29enterprise class routing protocols but
01:33here in the encore course cisco doesn't
01:36really have anything else about eigrp we
01:39just need to know primarily about OSPF
01:40except there is an item on the blueprint
01:43where we need to compare the
01:44characteristics of OSPF with a GRP
01:47you'll get more into eigrp if you take
01:49the an RC exam and study for that that's
01:51one of the concentration exams as part
01:54of the ccmp Enterprise cert track but
01:57here we just want to do a high-level
01:58comparison between OSPF and EIGRP
02:01and we're going to compare based on the
02:03following criteria we're going to
02:04compare the category of routing protocol
02:06that each protocol falls into will
02:08compare their administrative distance
02:10see which one is more believable
02:11than the other we'll talk about how each
02:13calculates their metric timers weren't
02:15different will distinguish between OSPF
02:18and EIGRP timers and finally we'll talk
02:20about how they differ in terms of load
02:22balancing but let's begin by
02:24understanding different categories of
02:26routing protocols and seeing where your
02:28OSPF and EIGRP fit in first let's
02:31consider these three different
02:33categories
02:33distance-vector link state and path
02:37vector the distance vector tells us in
02:40which direction do we go to get to a
02:43destination Network and how far is it if
02:46you think about your high school physics
02:48class or maybe in college and you
02:50learned about a vector remember a vector
02:52gives us two pieces of information it
02:54gives us magnitude and direction in
02:57other words go this way and here's how
02:59far it is that's what a distance vector
03:01routing protocol does it doesn't know
03:03the entire path that we have to go it
03:06just knows the next hop is in this
03:09direction we're gonna go to this next
03:10hop IP address or we're gonna go out of
03:12this egress interface and here's a
03:15metric or a cost representing how far it
03:17is to get to that destination network
03:20and a link state routing protocol is
03:23going to have a map of the entire
03:25topology the link state routing
03:26protocols use an algorithm it's called
03:28the Dijkstra algorithm it's the same
03:31algorithm that works on our navigation
03:32systems and our cars or if you have
03:34Google Maps on your smartphone those
03:36navigation systems that give you
03:38directions they have a map of the entire
03:41topology that you're gonna be traversing
03:43and they are going to assign a cost to
03:46each path and it's going to
03:49mathematically determine the lowest
03:51sequence of next hops that would get you
03:54there with the least cost so link state
03:57routing protocols don't just say go in
03:59this direction and here's how far it is
04:00it can look at the entire map of the
04:03topology and all the routers within this
04:05area of the topology they've got the
04:07same map so they all agree on how to get
04:09from point A to point B and a path
04:12vector and we're gonna see that only
04:13applies to BGP a path vector is going to
04:16say yes here's the direction that you go
04:18and no we don't have a map of the entire
04:20topology but we do know the paths we
04:23know the autonomous system
04:25that we have to transit to get to the
04:27destination Network now let's put these
04:29different protocols into categories rip
04:31the routing information protocol that's
04:33a distance vector routing protocol it
04:35says go out of this egress interface or
04:37go to this next hop IP address and
04:39here's how far it is here's the hop
04:42count rip measures how many routers we
04:44have to traverse our hop through in
04:47order to get to the destination network
04:49and by the way 15 is the maximum with
04:52rip if something is 16 router hops away
04:54it's considered to be unreachable
04:56now OSPF the main focus of this encore
05:00course is a link state routing protocol
05:02it's got a map of the entire topology
05:05and is is intermediate system to
05:07intermediate system that's also
05:10considered to be a link state routing
05:11protocol and it also uses that Dijkstra
05:14algorithm now EIGRP that's what we're
05:16comparing the OSPF with here in this
05:18video
05:19yeah GRP the enhanced interior gateway
05:22routing protocol it does fall under the
05:24distance vector category but it's an
05:26advanced distance vector routing
05:28protocol and what I mean by advanced is
05:30it's not as fundamentals rip just saying
05:32here's the best path and here's the how
05:34far it is and here's where I'm gonna go
05:36next
05:36no yeah ARP it can prequalify one or
05:40more alternate routes yes it'll find the
05:43best route and we'll talk a little bit
05:45about how it does that in this video but
05:47if that best route were to fail we don't
05:50have to go through a big massive
05:51calculation in all cases to see all
05:53right is there another way to get there
05:55no it's almost like we have a hot
05:56standby route standing by that we can
05:58switch over to in a matter of just like
06:00a couple of seconds or so so it's an
06:01advanced distance vector routing
06:03protocol and BGP the border gateway
06:05protocol that's often thought of as the
06:07routing protocol of the Internet as
06:09different autonomous systems are
06:11interconnecting they're going to
06:13typically speak BGP between themselves
06:15and BGP keeps track of the autonomous
06:19systems we have to transit to get to a
06:21destination network but with OSPF the
06:23main protocol we're focusing on in this
06:25course I said it's gonna have a map of
06:27the topology I want you to think of that
06:29map much like a puzzle and that map is
06:33stored in something called the link
06:35state database now this is a puzzle one
06:37of my daughter's was working recently
06:39and ahead like 1500 pieces to it and as
06:41I saw it it reminded me a lot of OSPF
06:44and I said I've got to take a picture of
06:45this to show my OSPF students the reason
06:48I say it reminds me of OSPF is she was
06:50working on this with her her boyfriend
06:53at the time fiance now and they would
06:55help each other and they would say do
06:57you have Cinderella's slipper or
06:58whatever it was and they would share
07:00pieces back and forth and slowly over
07:03many days they put together this puzzle
07:04well it's the same thing with OSPF with
07:07open shortest path first we've got
07:09different routers that know about
07:10different parts of our topology
07:12different parts of the network they're
07:14connected to different parts of the
07:15network and they're going to share
07:17information this information is
07:19contained in LSAs link-state
07:21advertisements and these link-state
07:24advertisements are essentially like
07:26these puzzle pieces
07:27we share the lsas between all of our
07:29members and together our collective IQ
07:32if you will amongst all the routers in
07:34the network they can assemble this map
07:36of the topology much like different
07:39people can assemble a puzzle because
07:40they each have pieces of the puzzle when
07:43you put them all together everybody has
07:45the same view of the puzzle just like
07:47all the routers within an OSPF network
07:49they've got the same view of the network
07:50once everybody has the same complete
07:53link state database another comparison
07:56between erp and OSPF is administrative
07:59distance administrative distance is the
08:01believability of a routing source for
08:04example here r5 is being told by various
08:06sources how to get to network 10.1.1.10
08:10t4r one says this is the way to that
08:13network r2 says here's how to reach that
08:15network r3 says i'll get you there and
08:17r4 says here's your ticket to 10.1.1.10
08:20t4 who do we believe well it depends on
08:23the source of the routing information
08:25and each type of route information
08:28source has an Associated administrative
08:31distance value and the lower the value
08:33the more believable that source if we
08:35have a network that's directly connected
08:37to us well that's pretty believable we
08:39know that that network is connected to
08:41us we know how to get to that network
08:42because we're part of that network and
08:44that's the most believable of all so
08:46that has an administrative distance of a
08:48zero however we could have a static
08:51route that we admin
08:52strangely add to the router by default
08:54it's going to have an administrative
08:56distance of a1 so that's pretty
08:58believable but then we start to get into
09:00our dynamic routing protocols ERP has an
09:03administrative distance of a nanny and
09:05OSPF has an administrative distance of
09:08110 that means that ya GRP is going to
09:11be more believable than OSPF by the way
09:14this is for e add European in a network
09:17in other words it's within an autonomous
09:19system we're not redistributing another
09:21autonomous system into EIGRP if
09:24something gets redistributed into EIGRP
09:26that EIGRP route is gonna have an
09:28administrative distance of 170 but here
09:31we're assuming that everybody's part of
09:32the same network and we've got an
09:34administrative distance of a nanny now
09:36rip is less believable still with an ad
09:38of 120 so that's another comparison
09:41between OSPF and EIGRP specifically yeah
09:43ARP is a little bit more believable now
09:46let's consider how OSPF and EIGRP are
09:49going to calculate the best path to get
09:51to a destination Network OSPF is gonna
09:53use a metric of cost and the lowest cost
09:56path is gonna win and the cost is the
09:58reference main width divided by the
10:00interface being with and by default and
10:03we can change this but by default the
10:05reference beam width is 100 megabits per
10:08second and we're trying to get from pc1
10:10to pc2 on screen and one way we could go
10:14from r1 to r2 to r3 and then out to pc2
10:17or we could go from R 1 to R 3 and then
10:20out to PC 2 if we were running a routing
10:23protocol like rip which only takes into
10:25account hop count it doesn't pay any
10:27attention at all to bandwidth then it
10:30would say oh it's a shorter hop count to
10:32go from R 1 to R 3 at to PC 2 as
10:34compared to going from R 1 to R 2 to R 3
10:37up to PC 2 it's one fewer hop but we can
10:40look at this and say I don't want to use
10:41that path of or 1 2 or 3 because that's
10:44a 10 Meg link I'd rather stay on the
10:46hundred Meg links and OSPF is going to
10:48agree with us so let's see how OSPF
10:50would calculate this cost OSPF would say
10:52the reference beam width is 100 Meg the
10:55link speed is 100 Meg and 100 divided by
10:58100 is 1 so we can give a cost of 1 to
11:01each of our 100
11:03link's what about that bottom link from
11:05R 1 to R 3 well that's the reference
11:08bandwidth of 100 divided by the link
11:10speed of 10 that's gonna be a cost of 10
11:12so we can tell from this that we would
11:15rather go from r1 to r2 to r3 that's
11:19gonna be a total cost of 3 as compared
11:21to going from R 1 to R 3 and then at the
11:24PC 2 which would have a cost of 11 so
11:27here is our preferred path here's a word
11:30of caution though notice I said that the
11:33default reference bandwidth was 100
11:35megabits per second what if we had a gig
11:37link which is very common or a 10 gig
11:40link if we had a gig link and we had 100
11:43Meg divided by 1,000 Meg a gig what
11:47would the cost be would seem to be a
11:49decimal of 0.1 we're not allowed to have
11:53decimal values for our cost so what
11:55would we do we would round up to a 1 if
11:58we had a 10 gig link or a 100 gig link
12:01they would all be a decimal value and
12:04they would all round up to 1 that means
12:07by default from OSPF sper spective
12:10we've got the same cost for 800 mega as
12:12we do for a gig a 10 gig or even a even
12:15a hundred gig link that's not good what
12:18can we do well is the best practice I
12:20like to set my reference bandwidth to a
12:22higher value I like to set my reference
12:24beam width to something like 200
12:26gigabits per second that way those
12:28different high bandwidth links they're
12:29gonna have different cost values so just
12:31a real-world caution for you now let's
12:34compare this cost calculation and a path
12:37selection with a GRP ya GRP considers
12:40more than just being with and here's an
12:42acrostic cut to help you remember this
12:44remember the acrostic big dogs really
12:47like me
12:47big dogs really like me and the being
12:50big is going to remind us of the be in
12:52bane width that's one of the things that
12:54he had your pee-can consider the D in
12:56dogs or reminds us of the D in delay the
12:59are in really reminds us of the are in
13:01reliability the Ellen like reminds us of
13:05the L in load and the M and me reminds
13:08us of the M and M to you now let's talk
13:10about each one of these the bane width
13:12is really the weakest
13:14as we're going through all these routers
13:16to get to our destination whichever link
13:18is the slowest bandwidth that's the
13:22value that we're going to be using for
13:23the bandwidth in fact the way we
13:25calculate being with is we take 10 to
13:27the seventh power and we divide it by
13:29the link speed that's the slowest link
13:31from point A to point B the delay is
13:34cumulative it's not the weakest link
13:36every time we exit a router interface
13:37we're adding on a little bit of delay so
13:40delay is cumulative and we'll get more
13:42into this in the an RC course if you
13:44join us for that one reliability though
13:46it talks about how reliable the link is
13:48and it's some number over 255 if we're
13:52not dropping packets we're 100% reliable
13:55then that value is gonna be 255 the load
13:57describes how busy our link is even
14:01though a link might be a high-speed link
14:02if it's congested it's still not a good
14:05link for us so load is a number over 255
14:08where a minimally loaded link would have
14:11a value of a1 over 255 so we might have
14:14a load of 1 that would be a minimally
14:16loaded link and it's a tie breaker we
14:19could use an MTU the maximum
14:21transmission unit the default is 1500
14:24bytes for a packet but sometimes we
14:26reduce the MTU because we have some
14:28extra header associated with a tunnel or
14:30different reasons but the highest MTU
14:32can win as a tie breaker the MTU itself
14:35though does not go into the formula and
14:37the formula is a bit intimidating to be
14:40honest here it is here's the formula for
14:43the eigrp metric now I don't expect you
14:45to memorize this for the exam but I want
14:47you to understand that there are ways of
14:50influencing the metric to pay more
14:52attention to bane with or delay or
14:54reliability or something like that for
14:56example if I wanted to pay more
14:57attention to the beam with value I could
14:59change that k1 variable to a higher
15:02number by default that K 1 value is a 1
15:05in fact here are all of the default K
15:08values that we have with the edge ERP
15:09now this is with traditional ERP there's
15:13something now called name to ERP where
15:16we actually have 1/6 K value again we'll
15:19get into that in the NRC course this on
15:21core courses that focused on EIGRP but
15:23notice the default K values there are 3
15:27the matter zeroes k2 k4 k5 they're all
15:30zeros and if we plug all those in the
15:32formula honestly kind of breaks down
15:33look at that k5 you see the k5 is zero
15:37and part of the formula is k5 divided by
15:40K four plus the reliability
15:43well if k5 is zero that whole thing is
15:45zero and we're multiplying zero times
15:47everything in the parentheses it's all
15:48going to be zero
15:49well there's an exception of this it
15:52says that yeah this is a formula unless
15:55k5 is 0 if K 5 is zero then we assume
15:58that K 5 over K 4 plus reliability we
16:01assume that entity is a 1 huh but that's
16:06the way it works now if you do the
16:08plugging and chugging net you'll see
16:09that it really falls out to where Yad
16:11Europeans really only considering being
16:13with and delayed by default in fact that
16:15was a question I missed on a technical
16:17interview when I was getting a job down
16:19at Walt Disney World as a network
16:20designer they used AI GRP on over 500
16:23routers at Disney World and they asked
16:26me by default
16:27what does EIGRP used to determine the
16:31metric and I knew this acrostic so I
16:33rattled it off I thought big dogs really
16:34liked me and I said really quickly I
16:36said it's been with delay reliability
16:39load and him to you and they said no
16:41that's wrong
16:42and I thought what do you mean it's
16:43wrong but then they pointed out that
16:45they had asked by default by default
16:48we've got all these zero K values we're
16:50not using him to you as a tiebreaker by
16:52default we're only considering bein with
16:54and delay and that's how the metric is
16:56calculated now let's see how AIG RP will
16:59use that metric in making its past
17:01selection decision take R 3 for example
17:04let's say that our 3 wants to get to the
17:06network of 10.1.1.1 the left-hand side
17:09of your screen it can get there from r1
17:11or r2 it's got to e add européenne
17:14neighbors and they're each going to
17:17report their distance from the network
17:19their metric now each of those routers
17:22r1 and r2 their metric to get to
17:2510.1.1.10 at their metric is we'll say
17:291,000 ok they're gonna report that
17:32information to r3 r1 says I can get
17:34there in 1,000 and r2 says I can get
17:37there in 1,000 but r3 realize is that
17:40it's
17:41cost something to get to r1 it's gonna
17:43cause something to get to r2 so r3 is
17:46gonna say let me add those up r1
17:48reported a metric of 1000 to get to that
17:52Network that was the reported distance
17:55you see there on the column of Rd
17:57reported distance r1 reported that to me
18:00of 1000 but it cost me 10,000 because
18:03maybe I'm on a serial link maybe it's a
18:0510000 cost for me to get to r1 well I've
18:08gotta add those two up 10,000 plus 1000
18:11that's 11,000 that gives us what's
18:13called our feasible distance or the FD
18:16what if I went via r2 well it reported a
18:19thousand metric to me as well how much
18:22does it cost me to get to our to 5,000
18:25let's add those up 1,000 plus 5,000
18:28gives us a feasible distance of 6,000
18:31and I'm gonna go with the lowest
18:33feasible distance and that's gonna be
18:35the path via our - that's how a GRP
18:38calculates its metric and makes a path
18:40selection decision another distinction I
18:43want to make between OSPF and EIGRP in
18:45this video deals with timers when we
18:48talk about timers we're talking about a
18:50conversation that goes on between
18:51neighbors be they OSPF neighbors or EIT
18:54RP neighbors in order for one router to
18:57know that its neighbors still there it
18:59expects to hear a hello message from
19:01that neighbor periodically and the
19:04timers work a little bit different with
19:05OSPF and EIGRP
19:07with OSPF we have something called a
19:09Hello interval and there are default
19:12values for the Hello depending on what
19:13kind of network we're on and we can
19:15certainly set that but I want you to
19:17know about the Hello interval that says
19:19how long we as a router or gonna wait
19:22before we send another Hello message to
19:24our neighbor so we have a Hello interval
19:26and we've also got a dead interval that
19:29says how long am I going to wait to hear
19:30a Hello from my neighbor before I
19:32consider that neighbor to be no longer
19:34available and OSPF neighbors they have
19:37to have matching timers they have to
19:38agree on how often are we going to be
19:40sending these hellos to one another and
19:41how long are we going to wait before I
19:43consider you to be down compare that
19:46with ERP ERP has the same kind of hello
19:51interval it says how long are we as a
19:53router go
19:54to wait between sending our hello
19:56messages to our neighbor but here's
19:58where it gets different yeah our P does
20:01not have a dead interval it has
20:03something called a whole time and this
20:07says not how long we are going to wait
20:10to receive a hello from our neighbor
20:12this is information that we send to our
20:15neighbor we're telling our neighbor wait
20:18for me for this long if you don't hear
20:21from me for this duration of time
20:22consider me to be down so again the big
20:25difference is OSPF has the dead interval
20:28and it says if I don't hear from my
20:30neighbor within this time they're dead
20:32to me
20:32but with EIGRP I'm telling my neighbor
20:35hey wait for me for this long then you
20:39consider me to be down it's like when
20:41I'm going into a cave and I say if you
20:42haven't heard from me in 20 minutes send
20:44help or something we're telling our
20:46neighbor how long do it for us with
20:48eigrp that's the difference and finally
20:50let's compare how load balancing works
20:52between OSPF and EIGRP
20:54with OSPF we can load balanced across
20:57equal cost paths now in this example
20:59with this topology on-screen let's say
21:01that we're sending traffic from a device
21:03off of sw1 to a device off of sw3 well
21:07as we go down the topology and we get to
21:09our to our two's got a couple of ways of
21:11getting to sw3 it could go via r3 or it
21:14could go via our for now OSPF is going
21:17to consider only baned within its
21:18calculation and notice the link speeds
21:20they're different we have a 256 K link
21:24between r2 and r3 we've got a 512 K link
21:27between r2 and r4 so SPF is gonna say
21:30I'm gonna choose the best link I'm gonna
21:32go from R 2 to R 4 yeah I might use r3
21:34as a backup if r4 were to go down but
21:37I'm gonna just use our 4 it kind of
21:39seems like we're wasting bandwidth
21:41doesn't it I mean sure it's not as fast
21:43but I could still be sending some
21:44traffic via r3 I could send more v r4
21:47but it'd be nice if I could sense some
21:49via r3 as well o SPF doesn't do that but
21:53II add your pecan yeah ARP has a feature
21:55called the variance feature and the
21:57variance feature is a multiplier let me
22:00show you what I mean let's say that the
22:02feasible distance from the perspective
22:04of r2 to get to this network off of sw3
22:07is about
22:0810.5 million and if we go via our for
22:11the cost is about 5.5 million well what
22:15we can do in Europe you outer
22:17configuration mode is give the command
22:19variants followed by this multiplier and
22:22let's say it's variants - if we multiply
22:25our best feasible distance which is just
22:27over 5.5 million if we multiply that by
22:302 the variance multiplier we get about
22:3411 million what we're saying is now we
22:38will load balance across any link whose
22:41feasible distance is less than that 11
22:45point zero two eight nine nine two
22:47million not equal to but less than that
22:50value now think about that link of via
22:53r3 it's about ten point five million is
22:55it less than the 11 million number it
22:57sure is so even though it's not as good
23:01of a link as compared to the link going
23:02to r4 we can still use it we can still
23:05load bounce across it if we configure
23:07the variance feature and by the way if
23:09we do this
23:10yeah GRP is gonna be really intelligent
23:11about this and it's not gonna try to
23:13send an equal amount of traffic via r4
23:15and r3 it knows that r4 is better it's
23:18gonna give a proportionally higher
23:19amount of traffic to the link going to
23:22r4 as compared to the link going to or
23:24r3 but this lets us use what would be a
23:27just a dormant stained by link with OSPF
23:30and lets us actively use it to send
23:32traffic and that's a comparison between
23:34some of the key features of OSPF and
23:37EIGRP
23:39[Music]
23:47you
23:48[Music]