LSA types - question relating to FLG

Hi all, Re: Pages 251-253 of the ROUTE FLG....See attached image. "In area 0, only type 1 and type 3 LSAs exist" - Why is this? Shouldn't a type 5 LSA also exist in area 0 (coming from the ASBR) and show up on R1? Also - What LSA types are used to send link information from ABRs to ASBRs? Type 1 and 3? The book shows the flows of LSAs from the outside of the network IN, and within the network, but not really the inside of the network OUT. Thanks in advance!

OSPF-LSAs.jpg

Find more posts tagged with

Comments

Forsaken_GA

It's badly worded and explained. Long story short, the type 5's aren't really bound to an area. If you look at the ospf database on a router, your type 1 and type 3 will have an area notation, since these can differ by area, whereas the type 5's are notated without an area. Which makes sense, they're not actually *in* an area, the only pertinent information OSPF needs to know about a route advertised by a type 5 LSA is how to get to the ASBR that's advertising it, and that will be in a type 4.

So area 0 is propagating the type 5, and because of that, it'll be generating the type4 for the ASBR into area 1. Within area 0 there should only be type 1 and type 3's (no type 2's since there aren't any network types with a dr/bdr), and all OSPF routers will know about the type 5's.

I'm wondering if the text isn't referring to a different figure, however. That image has area 1 marked as a stub/totally stubby, so it should *not* be getting type 5 LSAs. The ABR should be injecting a type 3 that contains a default route instead. If that is the right image for the excerpted text, then the book made a big mistake.

wave

Thank you, that's very helpful. I'll have to do some more reading on type 5's.

I checked the book again and that's definitely the correct diagram. I actually read ahead a little to the section on totally stubby areas and I see the contradiction. They obviously botched that one up!

Cheers

VAHokie56

Wouldn't the wording of the diagram make more since if they had stated if it was a stub or a totally stubby area? If it was a stub then the paragraph should be correct right? but if it was totally stubby then it would be false...but seeing as how they decided to taunt us with "stub/totally stubby" its a bit cloudy

wave

VAHokie56 wrote: »

Wouldn't the wording of the diagram make more since if they had stated if it was a stub or a totally stubby area? If it was a stub then the paragraph should be correct right? but if it was totally stubby then it would be false...but seeing as how they decided to taunt us with "stub/totally stubby" its a bit cloudy

I also wonder why they chose to bring in a totally stubby area when the book doesn't cover the subject for another 20 pages or so.

Forsaken_GA

VAHokie56 wrote: »

Wouldn't the wording of the diagram make more since if they had stated if it was a stub or a totally stubby area? If it was a stub then the paragraph should be correct right? but if it was totally stubby then it would be false...but seeing as how they decided to taunt us with "stub/totally stubby" its a bit cloudy

No. Type 5's aren't flooded into a stub area. They say that area 1 has 5 different LSA types in it. The type 1's and type 3's are obvious, it would also have a type 4 for the route to the ASBR, plus the type 5's. Since there are ethernet segments, I'm guessing there are also type 2's.

The only way the text is correct is if area 1 is *not* a stub/totally stubby area.

Remember, the difference between a stub and a totally stubby area is that a stub area still gets inter-area routes via type 3 summary lsa's, but no external routes. A totally stubby area gets *no* type 3 LSA or external routes, except the one type 3 LSA default route that's injected via the ABR.

So if area 1 is a stub or totally stubby area, then the text is blatantly wrong.

wave

Here is the output of #sh ip ospf database - thought it might be helpful.

output.jpg

wave

Forsaken_GA wrote: »

... whereas the type 5's are notated without an area. Which makes sense, they're not actually *in* an area, the only pertinent information OSPF needs to know about a route advertised by a type 5 LSA is how to get to the ASBR that's advertising it, and that will be in a type 4.

So essentially all a Type 5 does is trigger the generation of a Type 4?

Edit: After further reading it seems that Type 1 LSAs from the ASBR to an ABR cause the ABR to generate a Type 4 LSA to tell its area routers how to reach the ASBR.

So the Type 5 is the injection of routes from another AS?

VAHokie56

Forsaken_GA wrote: »

No. Type 5's aren't flooded into a stub area. They say that area 1 has 5 different LSA types in it. The type 1's and type 3's are obvious, it would also have a type 4 for the route to the ASBR, plus the type 5's. Since there are ethernet segments, I'm guessing there are also type 2's.

The only way the text is correct is if area 1 is *not* a stub/totally stubby area.

Remember, the difference between a stub and a totally stubby area is that a stub area still gets inter-area routes via type 3 summary lsa's, but no external routes. A totally stubby area gets *no* type 3 LSA or external routes, except the one type 3 LSA default route that's injected via the ABR.

So if area 1 is a stub or totally stubby area, then the text is blatantly wrong.

I know type 5 does not flood into stub areas but I don't think that's what the paragraph is saying..at least when I read it.

the paragraph says "The type 5 LSA is flooded into all areas by default. The ABR for area 1, Router R1, creates a type 4 LSA describing how to reach the ASBR, Router BBR2"...that sounds like a stub to me in area 1 since router 1 is its ABR ( one of the at least)

"The type 5 LSA is flooded into all areas by default" this is true because all OSPF areas by default ( before any additional config ) are standard areas and will receive the type 5's

Its worded funny but I think it makes since

Forsaken_GA

VAHokie56 wrote: »

I know type 5 does not flood into stub areas but I don't think that's what the paragraph is saying..at least when I read it.

the paragraph says "The type 5 LSA is flooded into all areas by default. The ABR for area 1, Router R1, creates a type 4 LSA describing how to reach the ASBR, Router BBR2"...that sounds like a stub to me in area 1 since router 1 is its ABR ( one of the at least)

"The type 5 LSA is flooded into all areas by default" this is true because all OSPF areas by default ( before any additional config ) are standard areas and will receive the type 5's

Its worded funny but I think it makes since

Keep reading. It says that R1, as the ABR for area 1 generates a type 4 to reach the ASBR.

The same text also says only type 1's and type 3's are in area 0... so logically it follows that the type4 is going into area 1. But if area 1 is a stub area as the diagram indicates, it shouldn't get a type 4, because it shouldn't have any type 5's. The follow up picks of the ospf database that the book describes shows that area 1 does in fact have type 4's

Long story short, the most likely case is that the diagram is in the wrong place, or wrongly labelled

Forsaken_GA

wave wrote: »

So essentially all a Type 5 does is trigger the generation of a Type 4?

Edit: After further reading it seems that Type 1 LSAs from the ASBR to an ABR cause the ABR to generate a Type 4 LSA to tell its area routers how to reach the ASBR.

So the Type 5 is the injection of routes from another AS?

No, a type5 describes an external route. It is an actual route advertisement for an external route, and the route it contains is valid. However, the only information it has besides the route is which ASBR actually advertised it. The router may or may not have a route to the ASBR. If it doesn't, then it can't use the external route, because it won't have a valid route to the next hop. That's what the type 4 takes care of. It makes sure that areas have next-hop information for the type 5's by making sure they have a route to the ASBR.

VAHokie56

Forsaken_GA wrote: »

Long story short, the most likely case is that the diagram is in the wrong place, or wrongly labelled

Yeap...It happens sometimes