OSPF LSA Types Explained

I love this lesson.
U r the best

Hi August. As far as I know, when you redistribute into ospf you can specify the external type (E1 or E2 - E2 is the default).

E2 will have a default cost of 20 and, when advertised to other routers inside ospf , the cost will keep in 20, so type E2 dose not consider the additional cost between other routers inside OSPF.

E1 will add the cost when advertised inside ospf. So, if you have 3 hops between the ASBR and other router with links running at 100 Mbps, you will have to add 3 to the cost, so you will end up with a cost of 23 at the far end.

Jaime,
Assuming someone hasn’t monkeyed around with OSPF’s reference bandwidth, you are exactly right.

19 posts were merged into an existing topic: OSPF LSA Types Explained

Hi Rene,

Why we need LSA type 4 in OSPF? any specific reason we are using this.

Regards,
Ajay

Hello Ajay!

The LSA type 4 is used when you have multiple areas as well as an Autonomous System Border Router (ASBR).

First of all, an ASBR is a router that has at least one interface in an OSPF area (area 0 or otherwise) and at least one interface in another Autonomous System (i.e. a region of your network running another routing protocol, static routing or another autonomous OSPF implementation). The ASBR sends type 5 LSAs (external routes) to its OSPF area(s). The type 5 LSAs are flooded to all areas. However, they do not contain the necessary information to FIND the ASBR. The ABRs (which connect OSPF areas) which receive these type 5 LSAs will also send out a type 4 LSA to all other areas to inform all routers of the location of the ASBR. (This is quite well depicted in Rene’s original post with diagrams).

So, the type 4 LSA informs all other areas of the location (or the IP address) of the ASBR so they know the path that must be taken to reach the external routes advertised by the type 5 LSAs.

I hope this has been helpful!

Laz

Hi Rene,

In the below output, can you explain the output seen under “Net Link States”?

R1#show ip ospf database 

            OSPF Router with ID (1.1.1.1) (Process ID 1)

                Router Link States (Area 0)

Link ID         ADV Router      Age         Seq#       Checksum Link count
1.1.1.1         1.1.1.1         30          0x80000003 0x004CD9 2
2.2.2.2         2.2.2.2         31          0x80000002 0x0048E9 1

                Net Link States (Area 0)

Link ID         ADV Router      Age         Seq#       Checksum
192.168.12.2    2.2.2.2         31          0x80000001 0x008F1F

Hello Ananth.

The information under the Net Link States (Area 0) show the DR(s) in the OSPF Area 0. The DR is the router with router ID 2.2.2.2 and IP address 192.168.12.2. For more information, take a look at this post: https://supportforums.cisco.com/document/133976/reading-and-understanding-ospf-database

I hope this has been helpful!

Laz

Rene , one question?

When you say in LSA type 5 , do you mean a route that is being redistribuited into RIP , and then into OSPF ? because Type 5 is External , and 4 mean when another domain as RIP sends traffic to our ASBR.

thanks in advance…!!!

Rayniero,
You are right that a Type 5 LSA means an external route. This corresponds with any route learned by redistribution. It could be from another routing protocol (RIP, EIGRP, IS-IS, BGP, ODR, etc), or it could be from a redistributing static routes, connected routes, etc.

A Type 4 LSA does correspond with an ASBR, but not necessarily about traffic being sent to it. The Type 4 LSA tells OSPF routers who the ASBRs are and how to get to them. It doesn’t carry information external routes flowing from those ASBRs.

Hi Rene,

Thank you so much for you lessons.
I complete my CCNA in April and i want to start CCNP but first i want to gain experience.
I work in Network but more LAN so even if i think its better gain experience, i think i need to focus in study with book or in this site or i never will get the knowledge to get the CCNP.
What do you think is the better aproach to CCNP?
I agre with you to do first the ROUTE, my question is there are a minimum time of experience (Labor) to aim to CCNP or i can start right way?
I have 2 years of academic ,netacad and 10 months of working in area.
Thanks for yourpassion in teaching, i definitely will stick you networklessons in my path.

Hello Bruno.

It’s great to hear that you too have a passion for networking.

I can tell you from my experience with CCNA and CCNP that waiting to gain experience in a working environment to be able to more successfully complete my CCNP was of limited value. This is because the CCNP topics are very broad. No matter what kind of job you have in networking, you will never cover the broad range of topics that exist in CCNP. My suggestion to you is to start studying immediately, do as many labs as possible, and when you’re ready, go for it.

I hope this has been helpful for you!

Laz

Hi Rene,

It would be really helpful if you post lessons about isis protocol.
your way of explanation is really simple to understand.

Dear Rene,

I am confused why we need OSPF Area 0 and why should all the other Areas be connected with Area 0? Can you please explain?

Thanks

Manami,
The short answer: To prevent loops when dealing with multi-area topologies.

The longer answer:
The OSPF RFC 3509 has a section that talks about this. As I am sure you know, OSPF is a link-state protocol, but that is only true for OSPF behavior within an area. When OSPF starts to deal with multi-area typologies, OSPF actual behaves like a distance vector protocol. Distance vector protocols can have their own set of problems, one of which is the “count to infinity” problem. This is where two routers use each other to get to a particular destination, so the traffic will just bounce back and forth until the TTL expires.

Because OSPF behaves like a distance vector protocol between different areas, this means that OSPF also inherits this distance vector problem. To fix this problem, OSPF (and other link-state protocols like IS-IS), have a backbone rule. It requires all inter-area routes to flow through the backbone where a “split-horizon” like behavior will occur. This means that the backbone won’t advertise a route it learned from a particular area back into that same area. This stops the count to infinity loop.

Check out the RFC link I included for more details.

1 Like

Hello,

I want to write CCNP Security 300-209 on September. As you know Cisco Book is not available for this exam. If you help me to get a good document to clear this exam.

Thanks.

If only there were such a document!

u rock!!

Per statement : “In this LSA we will find all the routers that are connected to the multi-access network, the DR and of course the prefix and subnet mask.”

Here Prefix means , IP address of DR which is connected in Switch - correct ?

Also in multi access network since we connect all routers via a switch: what kind of switches these are ? L2 or L3 ? Do we need to configure IP/VLANs anything on the Interface of switch that are connected to these routers ?

Hi Rene

when you say : We will find R1, R2 and the DR in the network LSA

So what exactly we find : is it Physical IP Address of R1/R2/DR OR loopback IP ?

Thankyou