Route Summarization: Mechanism To Optimize Routing Tables And Shorten Essay Length.

Steps for Efficient Network Summarization

1. Summarization also reffered to as Route aggregation is a mechanism used on layer 3 devices such as routers to reduce the size of the routing table entries with a view of grouping the networks together using a mask that is able to incorporate all the networks under consideration[1].

The following steps can efficiently be used to summarize group of networks.

Steps

• Identify the number of networks to be summarized
• Identify the binary power equivalent to the number of networks (which number raised to the power of 2)will give the exact number of networks.Incase the number of networks given is not in the multiples of 2 always go for the binary power with subsequent higher value and not the lower value.
• Subtract the binary power from the smallest CIDR value attached to the networks
• Attach the new found CIDR value to the smallest network among the ones given as the summarised address.

160.25.30.0/24

160.25.31.0/24

160.25.32.0/24

160.25.33.0/24

160.25.34.0/24

160.25.35.0/24

160.25.36.0/24

160.25.37.0/24

160.25.38.0/24

160.25.39.0/24

160.25.40.0/24 

Step 1:Total of networks given

11

Step 2:Binary power equivalent(x)that can equaite to the number of networks given

2^^x = 11

2^⁴ = 16 = 4

We shall go with 4 because 2^³ gives 8 witch is less than the number of networks which in this case is 11.

Step 3: Subtarct the answer (4)from the smallest CIDR value given in this case they are all the same as /24

/24-4 = /20

/20

Step 4:Attach the CIDR value (/24)to the smallest network from the ones given as the summarized address.

160.25.30.0/20

2. Route aggregation commonly referred to as Route summarization is a technique used by service providers to reduce the size of the routing table in a network in respect to the internet layer of the TCP/IP reference model.For example one may have a a private address such as the following

192.168.0.0/27,192.168.0.32/27,192.168.0.64/27,192.168.0.96/27,192.168.0.128/27,192.168.0.160/27,19.168.0.192/27 and 192.168.0.224/27 being summarized as 192.168.0.0/24.Multiple routes are combined into a single route such as 192.168.0.0/24 for advertisement projection.

Routers normally communicate with networks that remote through a routing protocol such as standard static or dynamic(Routing Information Protocol,Enhanced Interior Gateway and Open Shortest Path First) that contatins information about other devices on an entire IP network.

Summarization of route promotes collection of routes that are contigous from the routing table of a layer 3 device such as Router and Multilayer Switch as a single aggregated route.More routes must be adjacent for route aggregation to function. 

In  a medium to large enterprise networks,IP addresses are subdivided from one big network to smaller networks using subnet numbers through a process known as subnetting which can be done according according to the number of hosts or networks.An IP address such as 172.16.0.0/16 can be subnetted to produce 100 networks as follows:

172.16.0.0/23

172.16.2.0/23

172.16.4.0/23

172.16.6.0/23

The entire process will entirely produce a total of 128 because the bits required to produce 100 networks from the IP address given is 7 bits which translates to 2^⁷ = 128 networks.

Route Summarization's Benefits and Application

Route summarization negates this process where masks are shortened to form a larger network address that can incorporate the others when communication in an IP network through a process known as supernetting.

Networks are expected to grow for a given organisation or institution and this directly affects the other existing networks within the IP route table where other networks are also accomodated.The packet size increases making routers not to effectively handle the large number of networks.This greatlly inceases the processing power and lowers the response time making the router not be able to handle large networks effectively.Lower number of entries in the routing table greatly inceases te rsponse time and makes the router work efficiently[1].

Data transfer through bandwidth is allowed with the route aggregation mechanism where many routes will have a representation one line in update packet.

In a routing table packaet size is significantly expanded based on what the router needs to advertise for example if the router advertises 1000 routes,1000 specific lines will be required in update packet.When the packets increase in this case,the number of lines required in a similar way also increases making the packet size to expand as well as the bandwidth used.Actual data transfer in this case may get less bandwidth abvailability.

More bandwidth for data transfer can efficiently be achieved through route summarization because multiple routes are advertised with one line of update packet which significantly reduces the size of the packet or datagram.

When data is encapsulated and flows through the layer 3 device,it must associate itself with an interface where traffic must be channeled to.Routing table lookup is perfomed which takes a considerable huge amount of time for response to be processed.

We can therefore conclude thet route summarization can greatly reduce the problem by not only minimizing the amount of time required to perform lookup but also the number of CPU cycles. 

3. The way devices on the aggregation layer are not expected to expected to know internal details.

Border gateway protocol uses a given number of prefix in this case for router A,B,C & D  from being adverstised to neighbors.The prefix which should be atleast one in the aggregation layer of the routing BGP table should have the Aggregator and atomic_aggregate attribute to be able to gatjer information such as the Router_ID and AS_path so as to let other routers know that are part of a given attribute.Aggregate statement of static route to Null is created by Border gateway protocol so as to prevent loops occurrence[2].

The Use of Route Aggregation on Service Providers and Corporate Networks

High-End performance devices reside in the backbone and are expected to know and process internal route details of the aggregation layer.

For internal networks on both backbone and aggregation layer Open shortest path first (OSPF)routing protocol is used to find the shortest path that is availbale between the routers (AtoE)[3].Border gateway protocol in this regard is more scable but slower to failto a new route.The structure of OSPF is hierarchical with the concept of areas usage like the backbone area being 0 for for both backbone and aggregation layer.

There must be a provision to install appropriate route backups between the two ASes to improve reliability.

The provision can be accomplished by way of forwarding data elements destined for a failed link through backup path wil the sole objective of preventing attribute in the packet.Nodes on a public network communicate using a number of routing protocols such as BGP which is used to route data between autonomous systems that share a unified policy.Transport layer protocol known as TCP allows inter-autonomous system of routing to ensure relaibility. 

Border Gateway Protocol is hugely considered to be the main routing protocol used on the public nework.In the context of  Multihomed envronment,the two service providers SYD and MEL can be connected from AS 100 where the customer is situated to the IT infrastucture through External Border Gateway Protocol. Peering agreements between BGP is required inorder  to permit transfer from N1 through SYD link and N2 through MEL link.The informtion exchanged between the two routers is known as BGP peers which is exchaged through BGP sessions exhibited through trasnport layer protocol known as TCP known for its reliability and connected oriented nature as compared to UDP.

Routing information transfer between SYD and MEL who are the ISPs will entail some terms and conditions for traffic exchange.BGP which is the routing protcol of choice evalutes the routing routing table and any information that can be found on multiple routes between SYD and MEL routers  in relation to N1 and N2.

The best routes destined to reach N1 and N2 through SYD and MEL in the routing table is determined by BGP Path selection algorithm. This is specified using a set of command and attributes such as choosing the route with the higest weight and if its not set then secting the one with the highest local preference.Traffic destined to N2 from N1 best path might not always be the shortest path from source to destinations as compared to Interior Gateway Protocol (IGRP) such as Open shortest path first and Enhanced Interior gateway load balancing protocol.

Role of OSPF and BGP in Summarization and Optimization

When using Border gateway protocol,networks are normally adverstised through the network statements by default configured through the IGP using the i origin and those undertaken through redistributed networks are taken as incomplete.IGP origin takes precedence over the EGP origin.When two values are presented the higher one will be preferred over the one that is viewed to be of lower value.The local preference default value is always 100 which is used during best-path calculation when iBGP peers make advertisements in a network. From the Network access point,iBGP peers path selection can be influenced without necessarily having an effect on eBGP peers because local prefernce is advertised the autonomous system.

Autonomous system path length is influenced by the inbound traffic advertisements to other Autonomous systems.This can be influenced when a given number of attributes such as AS-PATH and MED are sent by the service providers.Conditional route injection trough BGP is another way inbound traffic can be influenced best prefix-matching behaviour that is the longest. For sites with upstream connections inbound traffic can be manipulated which may work sufficiently but with a number of disadvantages such as scalability.[4] Inbound traffic policy can be overriden making it difficult to tell or guide another autonomous system the way to route traffic to the source.To be able to affect inbound traffic the issue of policy  outbound should be enforced.

5. Route reflectors are the focal point for IBGP when using Border gateway protocol.They can be able to make a network converge among IBGP routes to fellow IBGP peers.The best path algorithm for BGP is selected using Router-id where the one with the lowest is considered.The originator ID is always substituted with the router ID if the router was received with the route reflector.In the scenario where there are several routers and one of the routers follows a certain rule and does not advertise to its internal neighbors prefix,route reflectors will be able to fix the problem.

The route reflectors will basically instruct that given router upo receipt of advertisement from an IBGP neitghbor in this regard other routers which are neighbors,to break the rule to the other members of the system.

Its important to check the the status of other client routers apart from cluster 1 and cluster 2.Ther should be no other external autonomous system so as to have the ioslated environment of the default rules[5].This will pave way other routers or clients from having alternative way of learning the route especially when client 1 and 2 are stooped from forwading traffic to the neighbors.

iBGP peering in a network can be reduced by the route reflector without change of the next hop.In the case of eBGP peering configuration can be done on directly connected interfaces as compared iBGP which is done through the loopback addresses.

Synchronization is performed by route reflectors between all the clients such as cluster 1 and cluster 2 so that the data path of the actual data can occur.

Route reflectors should always be positioned centrally in the context of the diagram given in order to curb sub-optimal routing.They may be peered with regular BGP speakers.Its good practice to have all non route reflectors peer with route reflectors.

References

[1]V. Fuller, T. Li, J. Yu, K. Varadhan,"Classless Inter-Domain Routing (CIDR)",an Address Assignment and Aggregation Strategy,1993

[2] Griffin,G.,Timothy,"An analysis of BGP convergence properties",volume 29,Oct 1999.

[3] D. Katz, K. Kompella, D. Yeung,"Traffic Engineering (TE) Extensions to OSPF Version 2",Sept 2003.

[4] Xiaomei Liu,Li Xiao,"Inbound Traffic Load Balancing in BGP Multi-homed Stub Networks",Aug 2008

[5] M. Caesar,J. Rexford," BGP routing policies in ISP networks",Nov 2005.