OSPF - what is advantage of creating multiple area

[maheshpula109]

I got to know that by creating the area, reduces the frequent LSA/LSU updates but I also know that between the area frequent LSA/LSU updates are reduced by summarization and within the area, the frequent LSA/LSU are controlled by DR & BDR. So what the advantage of creating multiple areas then? 

[Deepak Kumar]

Hi,

The division into areas allows routers in each area to maintain their own topological databases. This limits the size of the topological databases, and summary and external links ensure connectivity between areas and networks outside the autonomous system.

Routers in an area maintain a consistent topological database and any changes within the area have to be communicated to all devices. As the network grows, problems develop because the database is using too much memory or because changes are causing the processor to be overworked.

As the various databases increase in size and the calculations become increasingly frequent, the CPU utilization increases while the available memory decreases. This has the effect of increasing the network response time, as well as increasing congestion on the link. These effects may result in loss of connectivity, packet loss, or system hangs.

OSPF floods all link-state changes to all routers in an area. Routers with many neighbors have the most work to do when link-state changes occur. In general, any one router should have no more than 60 neighbors. (This theory seems a bit old because today we have some heavy routers those can handle the about 200 to 500 routes in the . But it is not recommended. 

Number of adjacent routers in an area: OSPF uses a CPU-intensive algorithm. The number of calculations that must be performed given n link-state packets is proportional to n log n. As a result, the larger and more unstable the area, the greater the likelihood for performance problems associated with routing protocol recalculation. Generally, an area should have no more than 50 routers. Areas that suffer with unstable links should be smaller.

This is not possible that all routers are connected on a single multiaccess link (for example, Ethernet) so there will no much benefits over the Ethernet Point to Point link.  Each multiaccess network has a designated router, which performs two main functions:

1. Originate network link advertisements on behalf of the network.

2. Establish adjacencies with all routing devices on the network, thus participating in the synchronizing of the link-state databases.

Regards,

Deepak Kumar

[Joseph W. Doherty]

In a nutshell, less control plane "work" on each OSPF router.

Traditionally, the algorithm of most concern has been Dijaski's algorithm, as it's processing time, I believe, can grow exponentially with size of the topology. One way to mitigate that, is to limit the size of a topology by using areas. (The current implementations of Cisco's OSPF also supports an optional feature, ispf [https://www.cisco.com/c/en/us/td/docs/ios/12_0s/feature/guide/ospfispf.html], which can also mitigate the impact of topology changes.)

BTW, I seen a mid size corporate OSPF network go into OSPF meltdown due to a flapping link. This network comprised both Cisco and non-Cisco routers. The non-Cisco router's didn't have the timer "backoffs" built into their OSPF implementation as does Cisco.

[Richard Burts]

There are multiple aspects of OSPF areas which provide advantages of implementing OSPF areas. I will mention 3 of them.

1) As Deepak mentions when a link changes state OSPF requires that the change be advertised to all OSPF routers in the area and each router must update its tables and calculate the convergence algorithm. When OSPF is implemented with a single area then a change on any router affects every router in OSPF. If you implement areas then a change on a router affects only routers in that area, and routers in other areas are more stable and do not need to process the update.

2) OSPF requires that every router in an area have the same knowledge of the routes in that area. So it is not possible to do summarization of routes within an area. Implementing OSPF with areas provides the possibility of doing summarization between areas.

3) A regular OSPF area must have knowledge of all routes (all internal routes from all areas, and all external routes). Implementing OSPF with areas provides the possibility of creating areas that do not have to know all of the routes. A stub OSPF area knows the all the internal OSPF routes but knows no external routes. It uses a default route to get to the backbone which knows all routes. A totally stubby OSPF area does not know any external routes and does not know routes from other OSPF areas. It uses a default route to get to the backbone which knows all routes. A not so stubby area is like a stub area in that it does not know of external routes from other areas. But it does allow redistribution to occur and to create external routes in the NSSA area. It uses a default route to get to the backbone which knows all routes. So these types of areas reduce the size of the OSPF tables, reduces the size of the IP routing table, and simplifies the routing logic of the router.

HTH

Rick

[maheshpula109]

Okay as per yours point no 1 - I want to ask how the other areas would be stable since the LSA3 would be propagating to them making them to run the SFP algorithm due to changes happened in another area. So didnt understood that whats the impact on network when creating areas ? since within the area - LSA 1 will be flowed and the same update outside the area - LSA 3(the update) would be flowing which makes both the routers run the SPF again which are in two different areas. So i am not able to find the convincing answers as whats the advantage of creating areas

[maheshpula109]

Okay as per yours point no 1 - I want to ask how the other areas would be stable since the LSA3 would be propagating to them making them to run the SFP algorithm due to changes happened in another area. So didn't understand that what's the impact on network/router when creating areas ? since within the area - LSA 1 will be flowed and the same update would be flown outside the area -as an LSA 3(the update) making both the routers run the SPF again which are in two different areas. So i am not able to find the convincing answers as whats the advantage of creating areas since the processing and load is same across both the routers which are present in different area. I hope you got my point what i am trying to ask.

[Jon Marshall · Accepted solution]

A full SPF calculation is only done on LSA types 1 and 2 and so is only done within an area. 

Between areas there may well need to a recalculation in terms of overall cost to a remote subnet ie. a subnet in a different area but this is not a full SPF calculation. 

Jon

[Richard Burts]

Thank you Jon. Your explanation is spot on. (so +5) Yes an LSA 3 does require a new calculation of cost but does not require running the SPF calculation. So routers in the other areas are more stable.

HTH

Rick