ospf part7: OSPF areas, router types, ospf routes & LSA types

Hi mate, I’m back once again. This time will dig deeper on OSPF.  We will discuss some of important concepts on OSPF which will be useful as we transition on CCNP-Route

 

OSPF Areas

• Why do we implement areas on ospf network and what will be the use of that?
• –> Here’s how it goes, when there is a network changes  let us say a new network  address was added on ospf topology with 50 routers, let us go back on ospf process, the local router will create its own LSA, will send hello to multicast address 224.0.0.5 or all ospf speaking routers and will expect for a hello message which has its own router-id and so on and so forth that it reaches the stage of asking for LSU. That seems to be not a complicated process however our goal in routing is that all routers can reach or ping each others even though there are 50 routers in the network. This scenario is feasible but this will lead to excessive CPU utilization and slower convergence. Why? because if there are 50 routers connected on that ospf network, it may require time for router LSAs to be propagated from routers near the originating router to router remote to the originating router.

 

Solution for that is implementing OSPF areas. How does it work and how does it help?

First let us define OSPF areas as a logical grouping of OSPF areas. Let us remember that the LSA that routers send on our discussion awhile ago is just router LSA. Router LSA is only sent on to ospf speaking routers within the same area. When we activate an ospf network, say for example when we type:

en
conf t
router ospf 100
network 10.0.10.0 0.0.0.255 area 0

There are three effects that will happen to the network:

1.) Activates all the interfaces that can be covered on the network statement- if we got interfaces  with ip addresses 10.0.0.1 , 10.0.0.2 and 10.0.0.9, all of those interfaces will be activated since they are covered in the network statement

2.) Advertises the network address of the activated interface

3.) Puts the activated interface on the area specified.

Since our goal is to avoid slow convergence and  high CPU utilization, router LSA is sent only on the area where the local router belongs. So the question that I guess that runs on your mind is how will the information or LSA be sent to routers on different areas if the router LSA is sent only to  the area where the local router belongs. The solution is that another LSA is created and advertised but we will go further to that but let me discuss first an important concept before we proceed to that.

OSPF Area types

1.) Regular area- any areas that is not area 0

2.) Backbone Area – area 0

Just take the definition as it is so that you will understand some other OSPF terminologies which we will discuss later.

OSPF router types

1.) Internal routers- routers that belong only to single area

2.) Area Border Routers (ABR) – routers that connect regular area to backbone area. The name itself defines that it is the border between regular and backbone area.

3) Backbone Router – routers with at least one interface which is connected to backbone area or area 0

4.) External routers – these are routers that are connected to internet, non-ospf router and other ospf AS system. Other OSPF AS system means that it runs OSPF as well but on the other company. Let us say it runs OSPF on AT&Tand the external router runs OSPF on Sprint.

 

 

Alright, let us try to identify the terminologies that we discussed before and let us identify those on the image above.

OSPF area types:

a.)Regular Area- area 15 and area 25; defined as area 0

b.) Backbone Area – of course no other than but area 0

 

OSPF router types:

a.)Internal routers-  defined as router with single area; those are R1, R3,R5,R6, R8 and R9 on the picture above

2.) Area Border router – defined as router that connects backbone area and regular area, those are  R2 and R7 on the picture above

3.) Backbone Router – define as router with at least one interface connected to area 0

So on the picture above, those are routers R2,R4, R5, R6, and R7

4.) External routers – let us say that R6 is connected to a router that only runs EIGRP, it will be removed on the list of internal routers but considered as external routers.

Be reminded that a router can be ABR and backbone router at the same time. Or a certain router on an OSPF network can have two classification of ospf router types.

Three types of OSPF routes:

1.) Inter- Area Routes-  routes within the different area; denoted by ‘O IA” on the routing table

2.) Intra-Area routes- route within the same area; denoted by ‘O” on the routing table

3.) External Routes – routes from outside ospf domain;denoted by ‘O E1 or OE2″ on the routing table

 

Alright! So I am now able to define those OSPF terminologies and hopefully you understood it by example. So we will be dealing back with the question awhile ago: if router LSA is only sent to routers on the same area then how will the LSA created by the originating router be advertised to routers on different area? The solution is

SUMMARY LSA

The scenario goes like this

1.) A network was added on R3 so it will update its LSA and will send hello to R1.

2.) R1 will save  R3’s LSA and will send hello back to R1 up until R1 and R3 is fully adjacent.

3.) R1 will advertise R3’s router LSA to R2

4.) R2 will save R3’ s router LSA advertised by R1 and will follow OSPF process up until R2, R1 and R3 are fully adjacent. Now, since R2 is the ABR, its task is to create summary LSA to advertise the router from area 15 to area 0. Be reminded that summary LSA is only create by ABR.

So if we will define SUMMARY LSA – it is the LSA created by ABR to advertise router LSA of a router on different area to another area.

 

The thing here is that on R4’s perspective, he only got one LSA from R1, R3 and R2 . Meaning to say the router LSAs which must be advertised by R1, R2 and R3  are viewed by R2 as just one LSA or summarized as SUMMARY LSA. So on R4’s perspective after full adjacency on all routers, R4 only got 4 neighbors instead of 9 routers.

We will continue the scenario and definitely I don’t have to elaborate other steps since it will be repetitive. The steps and scenario will be on the picture.

Alright! We already achieved Full adjacency on all routers, hopefully you understood the ospf terminologies, ospf areas, router types, ospf routes and the new LSA that we got.

Talk to you soon mate!

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

ospf part 6: how ospf works? ospf process & Hello content

Hey there! To learn a protocol, always asks how it works. Always remember, protocol is  a set of procedures or set of rules. In ospf, there are actually 7 process or steps to achieve neighbor adjacency. Let us first enumerate those processes.

1.) Down

2.) INIT

3) 2-way

4.) Exstart

5.) Exhange

6.) Loading

7.) Full

1.) DOWN- There will be only two reasons why an ospf network would go down

a.)When an interface is configured with an stafic ospf neighbor

• this will be topic which is covered on CCNP – route wherein instead that a hello message is sent to a multicast address, it will be sent to just unicast address

 

b.) When hello timer expires and then dead timer takes place.

For us to easily understand, those timers and multicast address that I am telling you. We will discuss first hello message content and its timer.

Hello Message Contents:

1.) Router-ID of local router- router-id of the router which will send hello message

2.) Area-ID- area id of the interface of the local router

3.) Address & Subnet Mask – address & subnet mask of the interface

4.) Authentication type- the password or passive interface configured on the router interface

5.) Hello & Dead Interval – there dyou go!- Hello and dead interval are check alive timers. Hello interval is a message sent by local router to check if neighbor routers are still alive. If neighbor routers did not response after  the default value of 10 seconds, it will wait for 40 seconds for it to declare that the neighbor is officially down. Please take note that 40 seconds is just the default value of ospf dead timer. The values can be also configured.

6.) Router interface priority – likewise I mentioned on the previous blog, priority is used on election of master-slave router. We will further discuss this as we  go on, but just to give you a hint, priority value is used who will be the first router to send its LSDB

7.) DR  & BDR information – let us make it simple, these can be compared to the master-slave routers. DR stands for designated router and BDR stands for back-up designated router. We will dig deeper on this on the next blog pages

8.) Router-ID of the local router’s neighbor – this appears on the neighbor table of the local router

Alright so those are the contents of the hello message which will be sent by the local router to the neighbor router and also the local router should expect those information to be sent by neighbor router to him.

Always remember mate, that there can be multiple protocols which also sends information like hello message or its counterpart. So, the question is how will a router knows that the hello message  is supposed to be sent to him.

On the image above, the question is how will neighbor router will know that it is that he is the recipient of the hello message? Also how will R3 would know that since EIGRP is configured to him, he is not recipient of the hello message? The message is sent with destination address of 224.0.0.5 which is a multicast address. I mean to say that the hello message is broadcast to the network but intended only to the group of routers that run OSPF protocol.

Let us say that R3 drops the packet since he is not the recipient of the hello message and R1, the neighbor router received the hello message of R2. What will happen next?

R1 will counter check the hello sent by R2. And the following should match:

1.) Area ID – if R1 is on area 0, R2 must be also on area 0

2.) Area type – we will discuss this further but just to give you a hint these are normal, stubby, not so stubby area., sounds very complex but we will discuss this as we go on

3.)Network address & Submet mask

4.) Hello & Dead intervals

5.) Authentication information

6.)Router-ID must be unique

***********************************************************************

OSPF STAGE 2: INIT

How we will transition from DOWN state to INIT ?

 

Here’s how it goes! Let play a scenario wherein R1 and R2 is trying to establish neighbor adjacency

Step1 ospf interface of R1 got activated,  R1 will create its own LSA, remember LSA? HEHE,  R1 will send hello message to its neighbor but since no neighbor routers got ospf activated, hello timer will expire and will proceed to dead timer.

Step2  ospf interface of R2 once got activated,R2 will create its own LSA, R2 will send hello to R1. (

Step3 R1 will counter check the hello message sent of R2

Step 4 once all the parameters are achieved, R1 will add R2 as its neighbor and then the OSPF  process will transition from DOWN to INIT, so let us define INIT stage

INIT STAGE – is when all hello parameters are and then the neighbor router is added on neighbor table of R1, it will remain on INIT stage until it received a hello with its own router-ID. Sound confusing? we will explain that further as we transition to 2-way

Step 5 – Since R1 already added R2 as its own neighbor, R1 will send hello message once again to R2 with hello message wherein the neighbor is R2. Since the hello message has R2’s router-id which will be sent to R2, therefore it is now the time to transition to 2-WAY

***********************************************************************

2-WAY –>  If a router received a hello from a neighbor and sees its own router-id

is included in the list of neighbor

Step 6

Let us have a recap from step 5, since R1 is on INIT state and sent hello to R2 with its neighbor router-id as hello content, R2 will add R1 as its neighbor router and will skip to 2-way, Step 6  would be R2 will send hello to R1 with its R1’s router-id as hello content then  R1 will transition to 2-WAY. Once both routers are on 2-way state, it will transition to EXSTART

***********************************************************************

EXSTART – the process of election on routers who will send its DBD

DBD – database descriptor – summary of router-ids of all LSAs inside LSDB

Step 7 Election will be on master-slave process

Master- highest router -id

Slave – lowest router-id

***********************************************************************

EXCHANGE – when we already figured out the master and the slave among routers.

Step 8 On our case, R2 will be the one who will send DBDs. R1 will just confirm that R1 & R2 share the same DBD as a reply ro R2

***********************************************************************

LOADING – it is the process of checking the LSDB of those routers involved to verify that they share the same LSDB

Step 9 After exchange of DBDs , if there are some neighbor information that they missed, master router will send a link state request which will lead to link state update and link state acknowledgment. Here’s how it goes!

Let’s say:

R2 ( master) : hi R1 here  is my ospf database

R1 ( slave) : Got it master!  I will try to counter-check and double check it!

Let’s say R2 has a new network configured to its interface and that network will be 10.0.21.0/24

R1 ( slave): R2 ( master ) , it seems like I don’t have yet the new network 10.0.21.0/24 configured to your interface, can I request to have that one . – THIS IS WHAT WE CALL LINK STATE REQUEST

R2( master) : Ohh oaah! I missed to inform you that, ok  no worries, I will provide you that network as LINK STATE UPDATE

R1( slave): Thank you so much master, I hope that you consider this one as LINK STATE ACKNOWLEGMENT to inform you that I already received the new neighbor information that you got.

***********************************************************************

FULL- we are now done with the final OSPF process wherein we have full neighbor adjacency. Loading state will be finished if all the routers have the same LSDB

***********************************************************************

Just a quick summary of OSPF process!

Oh yeah! Apologies if it is very lengthy  and boring but I hope that I was able to educate you today!

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

ospf 5: OSPF tables

Hi mate! We’re back again and we’re getting more serious. We have to grasp more on OSPF operation and hopefully I will be able to explain it on the simplest manner. Title above focus on OSPF tables. There are just three tables that we have to remember on OSPF namely: neighbor table, ip routing table and lastly Link State Database LSDB table.

A.) Neighbor table

Verifying command: show ip ospf neighbor

Do you happen to remember this ?

Yup, we have already encountered already the neighbor table. So we can conclude that the OSPF table includes the router-id of the neighbor routers, priority set on each neighbor which we will deal further as we go on, the OSPF stage which we will discuss as well  later and interface of the neighbor router where the local router is connected and the ip address itself. Too wordy right? Lemme discuss the OSPF neighbor content one by one.

1.) Neighbor router’s Router-id – the identifier of the neighbor router that also sends OSPF updates

2.) Neighbor router’s priority – this is used on the election of DR and BDR routers. We will go further on the classification of  ospf routers, it seems like there will be slave and master routers somewhat like that but I cannot provide more details about this without getting unto the basic of OSPF operation.

3.) OSPF stage – the word ” FULL” is the final stage on the OSPF stages on establishing neighbor adjacency. We will have a separate discussion about that

4.) The IP address of the neighbor router exiting interface. We just used the same network topology we used on the previous OSPF topic and as you can see below:

The squared ones are those the interface of the neighbor routers of R2 and also covers the IP addresses 201.52.30.1 and 192.168.14.1

5. Interface of the neighbor routers- these are source of the OSPF updates

B.) IP routing table

Verifying commands: show ip route, show ip route ospf

I hope that I was able to help you to read routing table. It’s just very easy as there was a legend to follow what protocols appearing below. “O” on the ip routing table represents the OSPF protocol. Below is a  sample of routing table of R2 on the network topology.

To show only OSPF on the routing table, use show ip route ospf

C.) Link State Database (LSDB) table

First, let us define Link- State, link-state is just one of the routing protocol according to algorithm. Algorithms that I am telling you are those like distance vector, link state and hybrid protocols. The reference to define its metrics is the cost. Cost has computation equivalent to reference bandwidth over point to point bandwidth. The usual reference bandwidth is 100 kbps. So cost = 100 kbps / link bandwidth.

Anyway, we forgot our goal which is to define link state. Link refers to the interface and so link state refers to the status of the interface. Therefore we conclude that LSDB is a database of the status of the interfaces. However, that is just the layman’s term explanation to define LSDB.

Below is an image that shows the content of LSDB.

Alright, another term rise that makes the study confusing but you don’t  have to worry about that.Lemme define them one by one.

LSDB – technically is a database of LSA’s or link state advertisement.

LSA – contains the information of the interface like network  address, metric, type of network address etc. These are the information advertised by the interface once the network command was activated.

So let us discuss the content of LSA. I indicate on the the white image above that there is option A and B. Those  are actually categories of link state. Link state categories are A.) what’s with the interface — network address, metrics, type of network address and B.) is there router connected? –> ip address of neighbor router and router-id of neighbor router.

Let us try to check these LSAs by examining our original network topology.

Let us check the LSA of R2

Category A for interface from R2 to R1 ( what’s with the interface)

Network address:201.52.30.0 /24

Metric:since we don’t know the metric, let us check the speed of the link, by running command ” sh interfaces se 1/0″ since it is interface se 1/0

We can see that is 1544 kbps so definitely the cost or the metric is 100/1.544 = 64.766 and we can round it down to 64

Metric:64

Type of Network address:

There area two types of network address that we have to define;

a.Stub network- these are loopback address network, network of which interface has only one or no router connected on the other end

b. Transit network- network which interface has two or more routers connected

Since the interface that we are talking about is serial 1/0  and since it is only connected with one router which is connected to a  switch, we conclude that it is a stub network.

Type of Network address:Stub network

Let us try to verify the LSA on LSDB using this commands:

show ip ospf database

Highlighted on the above image is link id 2.2.2.2 because on area 0 on router 2, we got three LSA’s namely 1.1.1.1, 2.2.2.2 and 3.3.3.3. Since we would like to check the interface of router 2, we need to further check the content of LSA of link-id 2.2.2.2 by verifying command: show ip ospf database router 2.2.2.2

Honestly, there should be 7 links  but I only highlighted above the interface that we checked its LSA content. So let us try to check ” what’s with the interface?”

We got type of network as stub network, network address of 201.52.30.0 255.255.255.0 and metric of 64. There dyou go! we are able to check one of link state categories which asks what’s with the interface?

Let us try to check the 2nd link state category which asks ” is there a router connected?”

Yes, there is a router connected, and let us check the interface of router 1 that links to router 2 and it is the serial 1/0 of R1. So I just would like to clarify here that on a certain LSA of a router, we will see as well the state of the interface of neighbor router that provides OSPF update.

We need to define the IP address of local interface , local interface means the interface of the original router that we are working with, which is R2. so it must be 201.52.30.2

IP add of local interface: 201.52.30.2 

Also, we need to verify the router-id of the neighbor router, in our case, it must be R1. So,

Router-id of neighbor router:1.1.1.1

Let us try to verify:

Oh yeah! That’s all for now for the OSPF tables. Talk to you later!

 

ospf part 4: ospf authentication

Hi there mate! This will be the last discussion on OSPF configuration. The rest of the OSPF topics will discuss how OSPF works and the content of Hello packet. The last topic that we had was configuring passive interfaces for security reasons. This time, we will configure OSPF with authentication using encrypted passwords. The need for configuring authentication is very simple, it is to provide security to active interfaces.  I mean active interface are those interfaces that receives and sends hello or OSPF updates. They cannot be a passive interface so authentication will be the best process to secure those interfaces.

Just come to imagine the scenario below wherein

Instead that the interface serial 1/0 of router 1 is connected to router2. It was connected to hacking router so it will lead to security breach.

So let us get back to our original topology.

Authentication SYNTAX:

en 
conf t
router ospf {process#} <— process # 
area {area#} authentication message digest < — we set authentication on the entire area on specific router
exit
int serial x/x
ip ospf message-digest-key 1 md5 {password} <— the authentication used is MD5 , you can further check on google how it works 

Note: 1.) We set password on each interface since the security breach may happen on each                     interface

2.) We can set multiple passwords on each interface: the syntax will follow:

 

int serial x/x
ip ospf message-digest-key 1 md5 {password}
ip ospf message-digest-key 2 md5 {password}

Very easy configuration right? So let us try to check the configuration on each router.
en
conf t
router ospf 100
area 0 authentication message-digest
exit
int se 1/0
ip ospf message-digest-key 1 md5 cisco

As we can see, opsf neighbor adjacency went down. It is for the reason that the neighbor router is not also configured with authentication and so they will not established neighbor relationship. Another way to verify: show ip ospf neighbor

So no neighbor relationship established. We can now continue with R2 authentication configuration:

@r2

en
conf t
router ospf 100
area 0 authentication message-digest
exit

As I go on, I can see that neighbor adjacency went down even without configuring the interfaces. It is for the reason that authentication was configured globally so it affects all the interfaces involved. We need to go by interfaces so that it would be more secured and to specify the password. How we will enter a locked door without a key? There should be a key. We need to specify the password / key-string. Below shows that no neighbor adjacency was established.

Let us continue the configuration:

int se 1/0
ip ospf message-digest-key 1 md5 cisco
int se 1/1
ip ospf message-digest-key 1 md5 cisco
exit

Below is what happened when I configured authentication on R2. We can see that serial 1/0 went up since it is connected to R1 and R1 has already md5 configuration. We can see on neighbor table that only R1 exists since R3 doesnt have md5 configuration.

Below is the neighbor table of R1, it shows that he can now see R2

Lastly, let us configure R3:
en
conf t
router ospf 100
area 0 authentication message-digest
exit
int se 1/1
ip ospf message-digest-key 1 md5 cisco
exit

Nothing much special with R3 configuration. We already discussed that serial 1/1 is possible to turn up since all routers has already md5 authentication and also router 2 appears on neighbor table. One thing that I can add to is that you alway see that term “from loading to FULL”. Those are stages of neighbor adjacency which we will discuss on the next topics.

I guess right now, you wonder how ospf works? We will discuss that on the next topic.

OSPF part 3: passive-interfaces

Hi mate! We are going to proceed with configuration of passive-interfaces. First, let me emphasize the need of passive-interfaces.  We configure passive-interfaces for security reasons. Imagine if an intruder tried to hack the network and the interface that it was connected is sending OSPF updates, so what will happen is the information that was advertised by neighbor routers will be shared to the hacking router. Let us check the scenario below:

R1 is connected to a switch, when hacking router connects to a  switch, then OSPF updates will be provided by neighbor routers to hacking routers. There dyou go, the network may be manipulated by hacking router. To avoid this, let us disable the ospf advertisement on interfaces which are not needed.

SYNTAX:

en
conf t
router ospf {process#}
passive-interface default
no passive interface ( interface which should be active)
exit

configuration on each routers

@r1

en
conf t
router ospf 100
passive-interface default
no passive-interface serial 1/0
exit

@r2

en
conf t
router ospf 100
passive-interface default
no passive-interface serial 1/0
no passive-interface serial 1/1
exit

@r3
en
conf t
router ospf 100
passive-interface default
no passive-interface serial 1/1
exit

 

Verification:

You would see that the arrows are pointing on passive-interfaces

 

In the event that the active interface became passive, it will not appear on  neighbor table. Below is the figure that proves that OSPF neighbor establishment has been configured smoothly. We will discuss neighbor table later as we go on.

We will check R2

In the event that the active interface became passive, it will not appear on  neighbor table. Below is the figure that proves that OSPF neighbor establishment has been configured smoothly. We will discuss neighbor table later as we go on.

Let’s check R3

In the event that the active interface became passive, it will not appear on  neighbor table. Below is the figure that proves that OSPF neighbor establishment has been configured smoothly. We will discuss neighbor table later as we go on.

Allright, that is just one way to establish security on OSPF network. Next topic is establishing network security on OSPF network using encrypted key but first lemme take my lunch mate!