As many of you have noticed, I’ve ramped up my review with beefy notes in nice pastel colors – my lame attempt to attract the ladies. Other than my wife, who reads looks at this blog once in a blue moon, all my readers have been mostly males. That’s all good. I’m sure you all enjoy the cool-in-the-eyes theme I’m trying to emulate. I was told ladies like pastel.

At this juncture, I’ve been studying BSCI for about 3.5 months now. I can honestly say that I have learned a ton already. Not mastered anything yet, though. But I expect that will come. It’s exciting to be at this stage because there is just so much to learn. I feel like a huge tree of knowlege just sprang up in front of me and I’m free to pick the fruits it bears. I just have to be careful though. There’s a saying: keep your friends close, but keep your enemies closer. I don’t really have a point to that. It just entered my mind as soon as I typed “saying”. But there’s another expression that comes up often, specially to those who achieve a higher degree of learning: “The more you know, the more you learn you don’t know”. Something to that effect. What I’m discovering as I gain deeper knowledge of routing protocols, metric calculations, etc., I find myself asking more and wanting to know more. When I began studies for this track, I decided I would stick with just the intermediate-level texts and materials to keep me on track of a scheduled and calculated study pace. Back in July, I pictured myself to be taking the BSCI exam by the end of this month. No way that’s going to happen. By the looks and feel of things, I’ll be happy if I can take it before the end of December.

Many CCNP candidates feel that the BSCI is the beast of all the 4 tracks. And I can see why that is. That’s also why I realized that sticking with the intermediate-level books can only hurt me. It is during this time that my curiosity about everything new that I’m learning is at it’s highest. My mind is always asking why and how whenever I learn something new. The books I’m using do not always satisfy. So I needed to consume more stuff of good quality. I picked up Jeff Doyle’s Routing TCP/IP, volumes I and II to fill that void. I had to, otherwise I would end up satisfying my hunger with crap. Allow me to analogize : Whenever I get hungry at work, I try to avoid going to the vending machine and buying all the unhealthy stuff they put in there. When I have them, I try to munch on some healthy nuts: cashews are my preferred ones. Sure they’re high in calories, but they are also proven to be high in fiber, omega-3s, and unsaturated fat (good fat). Not only that, they are filling and they suppress hunger so you don’t overeat later. So if given the choice to satisfy my hunger, twinkies or the healthy sfuff? They both satisfy your hunger but the effects are different.

I felt it was kind of the same with my learning. While learning all these new stuff is keeping me hungry, I need to satisfy that hunger with quality foods, before twinky-knowledge gets in first and I no longer want or am too full to consume the good ones.

A simple example: my BSCI book taught me that in order for a non-backbone OSPF area traffic to reach another area, it must be attached to a backbone area. That is area 0. I wondered why. So I googled it. Google said that if I don’t do that (attach a non-backbone area to the backbone area) other areas will be unreachable. Knowing that fact might be enough to get a correct answer on the BSCI exam. But then, reading some of Jeff Doyle’s teachings, he offers a more satisfying explanation: One of the positive arguments for OSPF – as a link state protocol – is that it has complete map of the entire network. This helps prevent routing loops, as opposed to a distance vector protocol where a routers knowledge of the network depends on what the next-hop router knows, which makes routing-loops more likely without careful administrative oversight. Another advantage with OSPF is the ability to segment a network into separate “areas” when it starts to get too big. That is good for easing some stress on the SPF calculations on the routers. As a result of the segmentation, routers in the same area get the over-all map of the network in that one area only. But don’t fret my bebes, OSPF has a mechanism in which information from one area is passed on to another area. An area will have a special router type that collects information for one area, and another separate database for information from another area. People like us call it ABR. The ABR connects two OSPF areas and maintains separate databases for each of those areas. It then passes along a summarized (and sometimes not) view of one area to another and versa vice. Essentially, each areas rely on the ABR to tell them what it knows about the other area. This is where the concept of link-state routers having the over-all map of the network sort of breaks down. In essence, this whole design of areas needing to find out information from it’s next-hop router, the ABR, about the network on the other side is a distance vector principle. And because distance-vector is prone to routing loops,  a loop-free inter-area topology can be assured by forcing all areas to only attach to one other area – essentially forming a hub-and-spoke topology between non-backbone areas and the backbone area. We  just like to call it area 0. For your FYI you can find this explanation in his article in networkworld. Or better yet, for a beefier explanaion,  it’s somewhere around page 382 of Routing TCP/IP, Volume I, Second Edition. You’d probably want to read from the beginning of that section to get the complete picture.

Anyway, where am I at in my studies? Well, I’ve nearly completed my first phase, minus IPv6. To be honest though, I almost don’t follow the different phases I set out to follow in the beginning. FYI, first phase was read all sections all the way through; second phase was re-read and write detailed notes; third phase was review – with emphasis on getting ready for exam. Right now, the first and second phases are completely intertwined and I’m hoping to start test preparation by November. I’m finishing up my notes on OSPF, with intentions to go back and hammer it in – because there’s just so much to know. I finished reading BGP but the write up will probably be equally massive if not more. First phase on multicast is also done. But I’m in the middle of re-reading and adding external readings on it.

There’s just so much to go through, I don’t know if I can get it done by December. We shall see, my friends. We’ll just have to see.

OSPF Special Area Types

• Standard Area
  • Areas that can accept intra-area, inter-area, and external routes.
  • In other words, the accept link updates, route summaries, and routes from other AS.
• Backbone (transit) Area
  • The central area to which all other areas connect.
  • Labeled as Area 0
  • Has all the properties of a standard area
• Stub Area
  • Does not accept routes belonging to other AS.
  • Theses areas do, however, have inter-area and intra-area routes within the same AS.
  • In order to reach the outside networks, the routers in the stub area use a default route which is injected into the area by the ABR.
  • A typical application of this area would be a situation in which a branch office does not need to know the routes to every other offices, but instead uses a default route to the central office where it can reach other destinations from there.
  • Cannot contain ASBRs (except when the ABR is also the ASBR).
• Totally Stubby Area
  • Only allows intra-area (within the same area) routes and default routes injected to the area.
  • In other words, it does not allow external autonomous system routes or summary routes from other areas.
  • It uses a default route to send packets to networks external to the area.
  • Cannot contain ASBRs (except when the ABR is also the ASBR)
• NSSA
  • Allows the flexibility of importing a few external routes into the area while still trying to retain the stub characteristic.
  • An example would be a router that is connected to an external AS which is running a different routing protocol – RIP, maybe. The router is now considered an ASBR. If that router is configured as an NSSA, the NSSA ASBR can start generating a special LSA type 7. These type-7 LSAs are flooded into the NSSA area. When the LSAs hit the NSSA ABR, they are converted to type-5 LSAs and are then flooded throughout the OSPF domain.

Characteristics that qualifies an area as stub or totally stubby:

• Single point of exit from the stub area. If there are multiple exits, one or more ABRs inject a default route into the stub area and suboptimal routing paths are acceptable.
• All OSPF routers inside the stub area, including ABRs, must be configured as stub routers. Recall that matching area flag is one of the condition for neighbor relationship. When the routers are configured as stub, all stub routers set a flag (the E-bit) in their Hello packets to zero.
• Virtual links cannot be configured within, nor transit, a stub area
• No ASBR is inside the stub area. ASBRs produce type 5 LSAs and type 5 LSAs cannot exist within a stub area.
• The area is not the backbone area (Area 0).

Configure Stub Areas

• Stub Areas reduce the size of the LSDB inside an area, which results in less memory used.
• Since the area also receives less routing updates, the SPF algorithm also runs less frequently
• Type 5 External network LSAs, such as those redistributed from other routing protocols into OSPF, are not permitted to flood into a stub area.
• A single default route (to 0.0.0.0) replaces many external routes.
• To configure, use the following router configuration command to all routers within the area:

area area-id stub

• area-id identifies the stub area. It can be a decimal value or a value in dotted-decimal format, like an IP address.

• By default, the ABR of a stub or totally stubby area advertises a default route with a cost of 1.
• To change the cost default route, use the following router configuration command:

area area-id default-cost cost

• The parameters are:
  • area-id - identifies the stub, totally stubby, or NSSA. Can be a decimal or dotted-decimal value.
  • cost – cost of the route summary. Values range from 0 – 16777215.

Figure 1: Example OSPF Stub Area Configuration

R3 Stub Configuration on Figure 1:

R3(config)#int fa0/0
R3(config-if)#ip address 192.168.1.2 255.255.255.0
R3(config)#int s1/0
R3(config-if)#ip address 192.168.2.1 255.255.255.0

!

R3(config)#router ospf 100
R3(config-router)#network 192.168.1.0 0.0.0.255 area 0
R3(config-router)#network 192.168.2.0 0.0.0.255 area 2
R3(config-router)#area 2 stub

R4 Stub Configuration on Figure 1:

R4(config)#int s1/0
R4(config-if)#ip address 192.168.2.2 255.255.255.0

!

R4(config)#router ospf 100
R4(config-router)#network 192.168.2.0 0.0.0.255 area 2
R4(config-router)#area 2 stub

• area 2 stub of the last line of each configuration defines the stub.
• Each router in the stub area must be configured with the area stub command in order to form neighborhood relationships.

Figure 2A and 2B: Routing Table for Standard Area (No Stub Area)

• In a standard OSPF area configuration, notice that in both routers R3 and R4, the route to the RIP network 10.10.10.0 is advertised as a Type 2 External route (E2)
• Any inter-area routes are advertised as IA.

Figure 3A nd 3B: Routing Table in a Stub Area

• After the area 2 stub command was configured on R3 and R4, notice that R3 (which is the ABR) still shows the route to the RIP network as a type 2 External route (E2). That is to be expected because at the ABR is where the filtering of the external route happens.
• Pay particular attention to the route designate with O*IA. Notice how the previous advertisement of the external route is replaced with this inter-area default route. The ABR filtered the type 5 LSA and injected a default route instead.

Figure 4A and 4B: OSPF Database on Router 4 Showing Standard and Stub Area Differences

• In the first figure above, you can see the OSPF database information of R4, configured in a standard OSPF area.
  • Notice the different kinds of LSAs advertised into it:
    • Type 1 – Router Link States.
    • Type 3 – Summary Net Link States.
    • Type 4 – Summary ASB Link States.
    • Type 5 – AS External Link LSA.
• The second figure, after area 2 stub command was configured show only the Type 1 LSAs and the Summary LSA for the default route (0.0.0.0).

Configure Totally Stubby Area

• A totally stubby area is cisco-proprietary.
• Whereas the stubby area blocks external type 5 LSAs, a totally stubby area blocks type 5 LSAs as well as Type 3 and Type 4 LSAs.
  • Totally stubby areas, therefore, only recognizes intra-area routes and the default route 0.0.0.0.
• The ABR injects the default summary link 0.0.0.0 into the totally stubby area.
  • Each router picks the closest ABR as a gateway to everything outside the area.

• To configure, use the router configuration command:

area area-id stub

• At the ABR only, configure:

area area-id stub no-summary

• The no-summary parameter stops summary LSAs (Type 3 and 4), in addition to external LSAs, from flooding into the totally stubby area.

Example

• Using figure 1 as the topology, the only difference with the configuration of a stub area and totally stubby area is the addition of the no-summary parameter from the area stub command on the ABR.

R3(config-router)#area 2 stub no-summary

• Examine the differences between the routing table R4 configured in a stubby area and a totally stubby area:

Figure 5A and 5B: R4′s Routing Table for Stubby Area and Totally Stubby Area

• In the first figure above the stub area ip route for R4 shows a default route (0.0.0.0) being advertised as well as and Inter-area (IA) route to a network in another area.
• The second figure, only the default route is advertised.
• Next, we compare the OSPF database for a stub area and totally stubby area:

Figure 6A and 6B: R4′s OSPF Database

• Notice on the last screenshot that, other than Type 1 or 2 LSA, there is only a single Type 3 LSA in the OSPF database.
  • This one, single, Type 3 LSA will be the only other LSA that will ever be advertised in a totally stubby area. And it is used to carry the default route information.

Configure NSSA (Not-So-Stubby Areas)

• A non-proprietary extension of the existing stub area feature that allows the injection of external routes in a limited fashion into the stub area.
• It is described in RFC 3101.
• The NSSA “bends” a rule of the stub area – the rule that says that there cannot be an ASBR inside of a stub area.
  • It allows an ASBR to be present in the stub, while at the same time performs an essential function of injecting a default route into the NSSA instead of the external routes advertised by ASBRs.
• Redistribution into an NSSA creates a special type of LSA known as Type 7. This only exists in an NSSA area.
  • An NSSA ASBR generates the Type 7 LSA, and an NSSA ABR translates it into a Type 5 LSA, which gets propagated into the OSPF domain.
• To configure an NSSA, use the router configuration command to all routers in the NSSA:

area area-id nssa [no-redistribution] [default-information-originate] [metric metric-value] [metric-type type-value] [no-summary]

• The parameters are:
  • area-id - A decimal or dotted-decimal value that identifies the NSSA.
  • no-redistribution - (Optional) Used when the router is an NSSA ABR and you want the redistribute command to import routes only into the standard areas, but not into the NSSA area.
  • default-information-originate – (Optional) Used to generate a type 7 default LSA into the NSSA area. The keyword takes effect only on an NSSA ABR or an NSSA ASBR.
  • metric metric-value – (Optional) Metric used for generating the default route. Values can be from 0 – 16777214.
  • metric-type type-value - (Optional) OSPF metric type for default routes. Could be one of two values:
    1. Type 1 external route
    2. Type 2 external route
  • no-summary – (Optional) Allows an area to be an NSSA but not have summary routes injected into it. Thus, the area is a totally stubby NSSA.

Figure 7:Example OSPF NSSA Configuration

R2 NSSA OSPF Configuration

R2(config)#router rip
R2(config-router)#redistribute ospf 100 metric 5
R2(config-router)#network 10.0.0.0
!
R2(config)#router ospf 100
R2(config-router)#redistribute rip subnets
R2(config-router)#network 192.168.10.0 0.0.0.255 area 1
R2(config-router)#default-metric 150
R2(config-router)#area 1 nssa

R3 NSSA OSPF Configuration

R3(config)#router ospf 100
R3(config-router)#summary-address 10.0.0.0 255.0.0.0
R3(config-router)#network 172.17.0.0 0.0.255.255 area 0
R3(config-router)#network 192.168.10.0 0.0.0.255 area 1
R3(config-router)#area 1 nssa default-information-originate

• In the example above R2 is the ASBR that redistributes RIP routes into area 1, the NSSA.
• R3 is the NSSA ABR
  • This router converts type 7 LSAs into type 5 LSAs for advertisement into backbone area 0
  • R3 is also configured to summarize (summary-address) the type 5 LSAs that original from the RIP network.
    • 10.0.0.0 networks are summarized to 10.0.0.0/8 and are advertised into area 0
  • To cause R3 to generate an O*N2 default route (O*N2 0.0.0.0/0) into the NSSA, the default-information-originate parameter is used on the area area-id nssa command on R3.

Figure 8: Shows Type 7 LSAs in R3′s OSPF Database

Figure 9: Shows the R4′s Summarized  Type 5 LSA (O E2), originated from the RIP network

Figure 10: R2′s Routing Table Showing the Default Route Originated by the default-information-originate Command

NSSA Totally Stubby Configuration

R3(config)#router ospf 100
R3(config-router)#summary-address 10.0.0.0 255.0.0.0
R3(config-router)#network 172.17.0.0 0.0.255.255 area 0
R3(config-router)#network 192.168.10.0 0.0.0.255 area 1
R3(config-router)#area 1 nssa no-summary

• The no-summary parameter works exactly the same as the totally stubby technique.
• A single default route replaces both inbound external (type 5) LSA ans summary (type 3 and 4) LSAs into the area.

Figure 11: Default Route advertised to R2

• The flooding of Type 3 and Type 4 LSAs into the NSSA is blocked. Instead a default route is injected into the NSSA totally stub area as a type 3 summary LSA.

Resources:

1. What Are OSPF Areas and Virtual Links?
2. OSPF Area Types – Packetlife.net
3. OSPF Not-So-Stubby Area (NSSA)
4. How Does OSPF Generate Default Routes?
5. How OSPF Injects a Default Route into a Normal Area
6. How OSPF Injects a Default Route into a Stub or Totally Stub Area
7. How OSPF Injects a Default Route into a Not So Stubby Area
8. RFC 3101 – The OSPF Not-So-Stubby Area (NSSA) Option

This entry is not an authoritative guide. These are merely notes and rehash of the primary text materials and resources that I use. For a thorough guide of the BSCI course, consider purchasing Building Scalable Cisco Internetworks (BSCI) (Authorized Self-Study Guide) (3rd Edition) by Diane Teare and Catherine Paquet, as well as following the links on the resources section of this entry.

OSPF Route Summarization

• Route summarization allows only consolidated routes to propagate into the backbone area (area 0)
  • Without it every specific-link LSA is propagated into the OSPF backbone and beyond.
• Type 3 Summary LSAs and Type 5 External LSAs are not summarized by default.

Two types of summarization:

• Inter-area route summarization
  • Occurs on ABRs and applies to routes from within each area.
  • Does not apply to external routes redistributed into OSPF
  • For efficient effect, network numbers within the area should be contiguous to create the least number of summarized addresses.
• External route summarization
  • Occurs on ASBRs
  • Applies to external routes redistributed into OSPF.
  • It is again important to ensure that summarized external addresses are contiguous to avoid problems with overlapping subnet ranges.
    • OSPF is a classless routing protocol, which means subnet mask information is carried along the route advertisement. RIPv1, however, could cause problems when OSPF is redistributed into that domain. VLSM information that is redistributed into RIPv1 is lost and static routes have to be configured in the RIPv1 domain.

Configure OSPF Route Summarization on an ABR

• Use the router configuration command:

area area-id range address-mask [advertise | not-advertise] [cost cost]

• Following is a description of the parameters:
  • area-id
    • Identifies the area whose networks is to be summarized

• • address
    • The summary address of the range of network addresses
  • mask
    • The subnet mask for the summary route
  • advertise
    • Optional parameter which tells the router to advertise the range and generate a type 3 summary LSA
  • not-advertise
    • Optional setting that suppresses type 3 summary LSA and hide component networks from other networks
  • cost
    • Optional setting that sets the metric or cost for this summary route. This cost is used by OSPF to determine the shortest path.
    • The value is anywhere between 0 to 16777215

• Example:

R1(config)#router ospf 100
R1(config-router)#network 172.16.32.1 0.0.0.0 area 1
R1(config-router)#network 172.16.96.1 0.0.0.0 area 0
R1(config-router)#area 0 range 172.16.32.0 255.255.224.0
R1(config-router)#area 1 range 172.16.96.0 255.255.224.0

• area 0 range 172.16.32.0 255.255.224.0 - this command summarizes the network 172.16.32.0 from area 0 into area 1. The networks 172.16.32.0 – 172.16.63.0 is summarized into 172.16.32.0/19 by the ABR R1.
• area 1 range 172.16.96.0 255.255.224.0 – this command summarizes the network 172.16.96.0 from area 1 into area 0. The networks 172.16.96.0 – 172.16.127.0 is summarized into 172.16.96.0/19 by the ABR R1.

Configuring OSPF Route Summarization on an ASBR

• Use the router configuration command:

summary-address ip-address mask [not-advertise] [tag tag]

• Following is the description of the parameters:

• • ip-address
    • The summary address designated for a range of address
  • mask
    • The subnet mask used for the summary route
  • not-advertise
    • Optional setting used to suppress routes that match the address/mask pair.
  • tag tag
    • Optional value that can be used as a “match” value to control redistribution via route maps.
• Example

R1(config)#router ospf 100
R1(config-router)#network 172.16.64.1 0.0.0.0 area 1
R1(config-router)#summary-address 172.16.32.0 255.255.224.0

• In this configuration, instead of flooding 32 external type 5 LSAs (subnets 172.16.32.0 – 172.16.63.0) into OSPF, only 1 summarized LSA type 5 LSA is flooded.

OSPF Default Route

• Default Routes injected into a normal area can be originated by any OSPF router.
  • However, by default, the OSPF router does not generate a default route into the OSPF domain.
  • default-information originate: used by OSPF router to generate a default route.
• Two ways to advertise a default route into a normal area:
  1. Advertise 0.0.0.0 into the OSPF domain, when the advertising router already has a default route.
  2. Advertise 0.0.0.0 into the OSPF regardless of whether the advertising router already has a default route.
     • This second method can be accomplished by adding the keyword always to the default-information originate command.

default-information orginate Router Command

default-information originate [always] [metric metric-value] [metric-type type-value] [route-map map-name]

• The parameters are:
  • always - (Optional) Specifies that OSPF always advertises the default route regardless of whether the router has a default route in the routing table.
  • metric metric-value – (Optional) Metric for generating a default route. If value is omitted and the default-metric router configuration is not configured, the default metric value is 1. *Note that Cisco documentations indicate that the default metric value is 10. But testing shows that it is 1.

• Example

R1 Configuration:

R1(config)#router ospf 100
R1(config-router)#network 10.1.1.1 0.0.0.0 area 0
R1(config-router)#default-information originate metric 10
!
R1(config)#ip route 0.0.0.0 0.0.0.0 198.1.1.2

R2 Configuration:

R2(config)#router ospf 100
R2(config-router)#network 10.2.1.1 0.0.0.0 area 0
R2(config-router)#default-information originate metric 100
!
R2(config)#ip route 0.0.0.0 0.0.0.0 198.2.1.2

• The optional metric parameter is used on R1 to prefer the default route to ISP A.
• The default route generated has a metric-type of E2 by default.
  • Because of this, the metric remains the same as it travels through the area. As a result, all routers prefer ISP A over ISP B.
• The default-information originate command causes the router to send a default route to all its OSPF neighbors.
  • Notice in the configuration of R1 and R2, the network command does not include the connection to the ISP routers. Therefore, a default route is not passed to the ISP routers.

Resources

1. OSPF Design Guide – OSPF and Route Summarization
2. Chris Bryant – Route Summarization and the OSPF Null Interface
3. How OSPF Injects a Default Route into a Normal Area

This entry is not an authoritative guide. These are merely notes and rehash of the primary text materials and resources that I use. For a thorough guide of the BSCI course, consider purchasing Building Scalable Cisco Internetworks (BSCI) (Authorized Self-Study Guide) (3rd Edition) by Diane Teare and Catherine Paquet, as well as following the links on the resources section of this entry.

I know the notes have been coming far and few in between. I apologize to those who count on my detailed notes to help them on their studies. However, things have been exponentially busy and taking detailed notes have been very hard to do. That doesn’t mean that studying has stopped. In fact, I just about finished reading the 600+ pages of the BSCI Study Guide. So now I’m going back to where I stopped taking notes and slowly add more of them here. They will not be as quick and often as I used to do but I still plan to keep them coming. I’ve relied very heavily on them during my last 2 tests and they worked effectively for me. So I’ll try my hardest to keep them detailed.

Here’s a brief outline of what’s keeping my busy these days:

• MRTG – I’ve been spending some good quality time with this very nice tool to monitor traffic loads on our routers. And the great thing about it is, it’s free. I’ve spent a good amount of time getting it setup on my Windows desktop as well as a linux setup dedicated for monitoring (more on this below).
• SNMP – In order to really capitalize on the power of MRTG, one has to know SNMP fairly well. In fact, MRTG doesn’t work if SNMP doesn’t run properly. At least that’s my experience. So getting SNMP to work on our routers required some time to research, read, learn and implement. Somehow, getting it to work on our pix was also a bit of a challenge. But it works
• Linux – I’m as newbie as newbie can get when it comes to Linux. So bringing up a Linux environment from scratch and getting MRTG to run on it was quite a bit of a fun challenge. I’m finally able to get one to run Xubuntu on an old 600Mhz Pentium III laptop with 256MB of RAM. Oh the nice thing about it: mrtg runs after rebooting the machine without doing anything. I dont know how it does it but I’ll surely find out soon enough when I break and fix the machine again over time. I’m also trying to test out different distros with as small a footprint as can be especially with the amount of resources I have (un)available. So far, I’ve tried SliTaz and DSL but haven’t used them enough to give an intelligible opinion of them. Recently I’ve also been reading up on least resource-intensive ways of running dynagen/dynamips on linux. And I must say, I’m pretty excited about trying it out. They say you can run a small installation of linux and have 8-10 routers running without killing your machine. That sounds awesome and can’t wait until I can get it to work on my home computer.
• Lastly and most importantly, did I mention that we are pregnant again? We are now 3 months in and the last couple of months have been especially hard. She has been tired all the time and was feeling sick for a good month or so. That also affected my studies because I had to take over most of her share of household duties such as cooking and baby-ing the little monster. But now she’s feeling better (’cause now she’s cooking again ). Which also means I get to study as close to my usual pace again. We’ll see how things develop.

As far as notes go, I’ll try to keep them coming, but if they don’t, you’ll know why. Or won’t.

OSPF Routers and LSA Types

OSPF Router Types

• Different OSPF router types control the type of traffic that go in and out of OSPF areas.
• When an area becomes too big, some of the following concerns become important:
  • Freqency of SPF calculations
  • Routing tables getting bigger
  • LSDBs also getting bigger.
• A solution to an increasing network is to implement a hierarchical area structure for the OSPF network. Some advantages of multiple OPSF areas are:
  • Reduced frequency of SPF calculation
  • Smaller routing tables
  • Reduced LSU overhead
• Here are the different router types:
  • Internal router – router’s whose interfaces are in the same area. Routers in the same area have the same LSDBs.
  • Backbone router – These routers sit on the perimeter of the backbone area (area 0) so it has at least one interface connected to area 0.
  • Area Border Router (ABR) -
    • Have interfaces attached to multiple areas.
    • It contains a separate LSDB for each area.
    • Route traffic destined for or arriving from other areas.
    • Exit points for the area, meaning that routing information destined for another area can get there through the ABR of that area.
    • Can summarize routing information.
  • Autonomous System Border Router
    • Have at least one interface attached to another autonomous system, such asa RIP network.
    • Perform route redistribution – a process of importing non-OSPF information to the OSPF network and vice versa.
• A router can be more than one router type.
• For each area that a router connects, it maintains a separate LSDB. Routers in the same area will have identical LSDBs for that area.
• An LSDB is synchronized between pairs of adjacent routers. On broadcast (LAN) networks, an LSDB is synchronized between the DROTHER.

OSPF LSA Types

Each LSA is a record that holds information for the database. As a whole, all these records make up the entire topology of an OPSF network.

Type 1: Router LSA

• A Type 1 LSA, or Router LSA is, flooded by each router in an area. A type 1 LSA describes the collective states of the router’s directly connected links (interfaces).
• Each of the router’s links (interfaces) is categorized into four diffrent link types as follows:

• *A stub network is a dead-end link that has only one router attached.
• For each of these links, there is a link data field that provides 32 bits of extra information.
  • For most link types this is the IP address of the associated router interface.
  • For stub network links, this link data field contains the subnet mask.
• Type 1 LSAs also indicates OSPF cost for each link, and whether the router is an ABR or ASBR.

Type 2: Network LSA

• Generated by the DR.
• Generated for every LAN (broadcast) or or NBMA transit network. An example of a transit network is an Ethernet LAN.
• The Type 2 LSA lists all the attached routers that make up the transit network, including the subnet mask of the link.
• Type 2 LSAs never cross the area boundary
• The link-state ID for a Network LSA is the IP address of the DR’s interface that advertised it.

Type 3: Network Summary LSA

• Sent by the ABR.
• A type 3 LSA advertises routes from one area into other areas in the OSPF autonomous system.
• When type 1 LSAs reach the ABR, the information from the type 1 LSAs are sent out by the ABR to other areas in the form of type 3 summary LSAs.
• By default, OSPF does not automatically summarize groups of contiguous subnets. It also does not summarize a network to its classful boundary.
• By default, a type 3 LSA is advertised into the backbone area for every subnet defined in the originating area.
• Manual summarization should be used to alleviate problems caused by significant flooding from too many networks being advertised.
• Summary LSAs do not, by default, contain summarized routes. Therefore all subnets in an area will be advertised, unless of course the network operator configures manual  summarization.

Type 4: ASBR Summary LSA

• A type 4 summary LSA is used to announce the presence of an ASBR. Therefore a type 4 summary LSA is only used when an ASBR exists within an area.
• It identifies the ASBR and provides a route to it.
• The link-state ID is the ASBR’s router ID.
• The ASBR sends a type 1 router LSA with a bit (known as the  external bit or e-bit) that identifies itself as and ASBR. When an ABR (that is identified with a border bit or b-bit in the router LSA) receives this type 1 LSA, it builds a type 4 LSA and floods it to the backbone or area 0.

Type 5: External LSA

• Describe routes to external OSPF autonomous systems.
• These are generated by the ASBR and are flooded to the entire autonomous system.
• The link-state ID is the external network number.
• Again, because summarization does not occur by default, the network operator should consider manual route summarization at the ASBR to prevent problems with over flooding.

OSPF LSDB & Routing Table

OSPF LSDB

The command show ip ospf database allows one to view the contents of the OSPF LSDB.

Router# show ip ospf database

OSPF Router with ID(192.168.1.11) (Process ID 1)

                 Router Link States(Area 0)

 Link ID           ADV Router        Age         Seq#       Checksum Link count

 192.168.1.8       192.168.1.8       1381      0x8000010D    0xEF60   2

 192.168.1.11      192.168.1.11      1460      0x800002FE    0xEB3D   4

 192.168.1.12      192.168.1.12      2027      0x80000090    0x875D   3

 192.168.1.27      192.168.1.27      1323      0x800001D6    0x12CC   3

                 Net Link States(Area 0)

 Link ID          ADV Router        Age         Seq#       Checksum

 172.16.1.27      192.168.1.27      1323      0x8000005B    0xA8EE

 172.17.1.11      192.168.1.11      1461      0x8000005B    0x7AC

                 Type-10 Opaque Link Area Link States (Area 0)

  Link ID         ADV Router        Age         Seq#       Checksum Opaque ID

 10.0.0.0         192.168.1.11      1461      0x800002C8    0x8483     0

 10.0.0.0         192.168.1.12      2027      0x80000080    0xF858     0

 10.0.0.0         192.168.1.27      1323      0x800001BC    0x919B     0

 10.0.0.1         192.168.1.11      1461      0x8000005E    0x5B43     1

The following explains the purpose of each column:

• Link ID – Identifies the Router ID number
• ADV Router – Identifies the advertising router ID. This is the source router of the LSA
• Age - The age of the Link state. The maximum is 3600 seconds (1 hour).
• Seq# – The link state sequence number. The sequence number starts at 0×80000001 and increments by one each time it is updated. This helps detect old and duplicate LSAs.
• Checksum – Ensures the reliable receipt of the LSA
• Link Count – Shows how many links are attached.
  • Used only on Type 1 Router LSAs.
  • The link count includes all point-to-point, transit, and stub links.
  • Point-to-point serial links count as 2
  • All others count as one.

Route Types in the Routing Table

Different designations describe the route types generated by OSPF:

• O - Indicates that the route comes from within the router’s area. These routes are advertised by router LSAs and network LSAs
• O IA – The “IA” stands for inter-area. It indicates that the routes come from networks outside the router’s area (but still within the same autonomous system.) This type of route is advertised by ABRs through summary LSAs.
• O E1 – External LSA type 1. Route costs are calculated by adding the external cost to the internal cost of each link. This type is useful when multiple ASBRs are advertising external routes to the same AS – it avoids suboptimal routing.
• O E2 - External LSA type 2. The route coast never change and it is always the cost of the external route.

OSPF LSDB Overload Protection

• OSPF LSDB overload protection can protect the routers from resource (CPU and memory) drains. An example of such an instance is a misconfiguration of routers that causes a redistribution of a a large number of prefixes, in turn generating excessive amount of LSAs that are generated.
• This feature is available with Cisco IOS Software Release 12.3(7)T and later, as well as some specific earlier releases.

max-lsa maximum-number [threshold-percentage] [warning-only] [ignore-time minutes] [ignore-count count-number] [reset-time minutes]

The parameters are as follows:

Changing the Cost Metric

The general formula used to calculate OSPF metric is 100Mbps/(bandwidth in Mbps).

For example:

1. A 64 kbps link has a metric of 1562:
   • 64kbps/1000kbps = 0.064 –> 100Mbps/0.064Mbps = 1562.5
2. A T1 link gets a metric of 64
   • 100Mbps / 1.544Mbps = 64.7

• The problem with that formula is that the maximum interface it can do is 100Mbps, which will yield a metric of 1.
• For interfaces faster than 100mbps, use the auto-cost-reference-bandwidth ref-bw command.
  • The ref-bw is any range between 1 to 4,294,967 in megabits per second. The default is 100.
• Also, remember to use the bandwidth value interface configuration command to accurately depict the correct interface bandwidth, in kilobits per second

• The ip ospf cost interface-cost configuration command to override the default cost. The interface-cost is an integer from 1 to 65,535.
  • The lower the number, the better (and more preferred) link.

Resources:

1. Link State Advertisement Formats
2. IP Routing Protocols Commands – show ip ospf…
3. OSPF E2 vs E1 Routes – Chris Bryant
4. OSPF Link State Database Overload Protection

This entry is not an authoritative guide. These are merely notes and rehash of the primary text materials and resources that I use. For a thorough guide of the BSCI course, consider purchasing Building Scalable Cisco Internetworks (BSCI) (Authorized Self-Study Guide) (3rd Edition) by Diane Teare and Catherine Paquet, as well as following the links on the resources section of this entry.