With the scale of Internet traffic rising, the more high performances requests the router needs. The central control plane based on the Forwarding and Control Elements Separation architecture router has a certain limitations so it is complicated to meet future network requests. Therefore, there is practical significance that the functioning of distributed routing protocol and the enhancement of the control plane's performances better meet the needs of future network. [1]
Design Consideration
1. DESIGN OF THE DISTRIBUTED OSPF PROTOCOL OF FACING FORCES ARCHITECTURE
In order to entire distributed the OSPF protocol between the CE-CE, the OSPF protocol is divided into three modules: Signal processing module, routing module and routing table management module, so that they run on different CE ,each CE works together to complete the full functions of OSPF protocols. The signal module is dependable for processing Hello Protocol packages, through the Hello protocol packet processed to create and maintain neighborhood.At the same time the signal processing module is also dependable for flooding and the synchronous link state database. Route calculation module uses the Shrotest Path Tree to calculate the route, when the external LSA shows that the network topology changed, route calculation module will start the SPT algorithm to calculate a new route to update the routing table. Routing table management module is dependable for routing table maintenance and management, and updating on all the FE forwarding information table.[1]
There are three main methods about operation of Distributed Routing Protocol: routing protocol functional distributing , network layer, packet parallel processing :
i) Routing protocol functional distributing
ii) Network layer
iii) Packet parallel processing
Results
Test platform for a distributed OSPF protocol of facing ForCES architecture is primarily composed by the CE, FE,SmartBits, SmartBits analysis host, Ethernet switch and FE Development Host). The host has installed MontaVista.SmartBits is a test mechanism, which has a number of integrated network test functions and our maintest equipment. Test platform is shown in Fig. 2.
Conclusion
The content of this text is the design for a distributed OSPF protocol operation in ForCES architecture .Combining the distributed routing protocol, two models for the distributed OSPF protocol in ForCES architecture. The results prove the distributed OSPF protocol in ForCES architecture's possibility and advantages which are compared to central.[1]
2.Extending OSPF for Large Scale MPLS Networks
Introduction
Open Shortest Path First (OSPF) is classify as an Interior Gateway Protocol (IGP). This means that it distributes routing in sequence among routers belonging to a single Autonomous System (AS). OSPF routes IP packets based exclusively on the destination IP address establish in the IP packet header without any supplementary encapsulation. OSPF is link state based and uses the Dijkstra algorithm to calculate the shortest path between any node in the network to every other node. A router relations OSPF Hello messages with its neighbors to inquire about their status.[2]
In this journal its use LSAs. There are several types of these LSAs. These LSAs are originally generated and propagated based on the network topology and configuration. An addition or change in the network status or topology may cause the generation of related LSAs to return that. In OSPF is when an LSA is generated by a router, the router has to flood this LSA over its adjacencies. When a router receives a new LSA the router has to: process this new LSA on the first link it receives it on, process duplicates of that LSA when or if received on other links, acknowledge all received LSAs on all links and flood the LSA over all links except the one it was received on first. This causes a huge weight on the CPU of routers with many adjacencies in a large scale network.[2]
Design Consideration
In this work the change of standard OSPF was imitation by adding a new Type of Service (TOS) in the Router LSA per link and a new sub-TLV in the Link TLV of the TE LSA for OSPF-TE. For stability and ease we use the analogy of the TOS in the Router LSA through out the paper. The added TOS was used to specify two criteria: link priority and flooding priority.
ThisTOS was used to indicate the precedence of each link for forwarding data or setting up an LSP over that link. In additional it was also utilized to specify whether to use that connection to flood control traffic such as LSAs over it or not to realize optimized flooding[2]
In addition, the journal illustrate how parallel links between two routers with related characterization(bandwidth, cost metric and usage) can be represented in OSPF using what we call a virtual trunk to reduce the number of LSAs being generated in an OSPF domain.[2]
In this journal recommend a conversion of OSPF to optimize its performance and therefore raise network reliability and scalability. The adjustment can be applied to both standard OSPF and OSPF-TE. The real meaning of our approach is in applying the adjustment to optimize the flooding of LSAs on parallel links and to load balance traffic over Equal Cost Multi Paths (ECMP) in a Multi Protocol Label Switching (MPLS) network running OSPF or OSPF-TE.[2]
Conclusion
Internal gateway protocols like OSPF are presently creature used in large scale ISP networks which are increasing fast in size due to increasing require for bandwidth. In this paper we have discussed and proposed new ways to optimize OSPF and OSPF-TE for large scale MPLS networks. The optimization expected at load comparison LSPs over parallel links and minimizing flooding of OSPF control traffic by extending the information provided through OSPF LSAs to help realize that. The simulation results presented here, showed how the future load comparison schemes can better realize link utilization regularity than standard OSPF, yet minimal changes were required to accomplish that. [2]
3. OSPF Failure Identification based on LSA Flooding Analysis
Introduction
It is important to monitor routing protocols for stable operation of IP networks. In this paper, its focus on the Open Shortest Path First (OSPF) , a widely deployed intra-domain routing protocol. Routers running OSPF advertise their link states on Link State Advertisements (LSAs) to the network, and OSPF routers can construct a complete view of the topology of the network from these LSAs. Therefore, by monitoring LSAs flooded throughout the network, we can comprehend the topology of the network. OSPF routers also advertise LSAs immediately when they detect the change of their link states caused by failures, etc., so the location of link failures on the IP network can be detected by monitoring LSAs[3]
Design Consideration
In this journal said that the work the modification of standard OSPF was simulated by adding a new Type of Service (TOS) in the Router LSA per link and a new sub-TLV in the Link TLV of the TE LSA for OSPF-TE. For consistency and simplicity, we use the analogy of the TOS in the Router LSA through out the paper. The added TOS was used to indicate two criteria: link priority and flooding priority. This TOS was used to indicate the priority of each link for forwarding data or setting up an LSP over that link. In previous it was also utilized to specify whether to use that link to flood control traffic such as LSAs over it or not to realize optimized flooding [6]. The value of this TOS can be set originally by the administrator when creating the link then restructured according to the dynamics of the network. In this paper assigned the TOS value based on the bandwidth utilization of the link, utilizing three schemes for load comparison, the main concern scheme, the risk priority scheme and the incremental priority scheme. For all schemes a path is chosen first based on the original cost metric and an LSP was only set if the necessary bandwidth is available, otherwise the LSP will be blocked.[2]
In this paper, its recommend a method of OSPF malfunction recognition that takes into account these two aspects. In OSPF, the pattern of LSA flooding when an OSPF malfunction happens is deterministic. Therefore, by analyzing the pattern of it in feature, its can recognize OSPF failures perfectly by passive monitoring of LSAs. However, since it is non-deterministic when LSAs are experimental at a monitoring point, the timer for waiting delayed LSAs should be carefully designed. [3]
These papers described that they can track changes in the network topology and detect failures of links or routers in the network by monitoring LSAs. However, no papers analyze the LSA flooding in detail, i.e., they did not mention the organization of multiple LSAs flooded when failures occur or LSA delay in detail.[3]
CONCLUSION
In this paper, its future a method of OSPF malfunction recognition based on the analysis of LSA flooding patterns. They described the algorithm of identifying OSPF failures based on the study of LSA flooding patterns, and the contemplation of organization of multiple LSAs and LSA delay in an OSPF network.
Then established their method on the experimental lab network and showed that it recognized some OSPF failures accurately They provide some discussion items about OSPF malfunction recognition and showed possible solutions. In expectations work, they plan to install their OSPF monitoring system to a real network, and estimate the performance and efficiency of our method.[3]
The main contribution of this paper is that recommend an algorithm for identifying OSPF failures such as failures of a router and a transit network based on the feature analysis of LSA flooding patterns, and the consideration of association of multiple LSAs and LSA delay in an OSPF network. Analyze the LSA flooding in detail, the method can also identify multiple OSPF failures almost correctly even if they arise simultaneously[3]
Result and discussion
From journal 1 the result are identify by2 stage :
i) Test platform
ii) Analysis of performance
Analyzing LSA Flooding and Failure Identification:[3]
1. Autonomous routerfailure
2) Accidental routerfailure
3) Transit networkfailure
4) PtoPlinkfailure
5) Transit network linkfailure
4.Practical OSPF Traffic Engineering
Introduction
MOST of today's traffic engineering (TE) proposals need the consumption of costly routing and traffic forwarding hardware and software. On the other hand, ISPs have huge installation base of routers running best-effort routing protocols, like Open Shortest Path First (OSPF) [2]). OSPF provides shortest-path-first routing, simple load balancing byEqual-Cost-MultiPath (ECMP: traffic is split roughly evenly amongst equal cost paths) and means to manipulate routing through setting the administrative link weights.
Hence, it is an easy-to-deploy and overly cost-effective solution to implement traffic engineering on top of OSPF, while retaining obtainable routing equipment. In such an architecture, a suitable Traffic Engineer: participates in OSPF signaling to learn routing information; assigns paths for each session; computes link weights as to assure that link weights reflect the assignment of paths (i.e., all paths, which are assigned for a exacting session are shortest paths for the session); and distributes the chosen link weights back to OSPF routers.
Technology
OSPF TE must work with or without precise knowledge on the capacity of network links. Additionally, a practical TE algorithm must-under all circumstances-afford practical link weights. Any unintentional interfering of the shortest paths of different sessions (ties) must be avoided. OSPF-acquiescent weights are integer esteemed and fall into the range .The link weight computation algorithm must run speedily to assure quick adjustment to topology changes or management controls. Naturally, it also must defer competent routes.
OSPF traffic engineering is, in general, NP hard, and propose a local search heuristic
algorithm achieving close to optimal performance'
Result
Our most main conclusion is that, given the normal improbability of ECMP load corresponding it is not worth precisely
provisioning OSPF routing w.r.t. a given demand matrix. Instead, our priority-driven comparative dimensioning scheme performs logically improved. In addition, correctly setting the preferred distances and using tie breaking may be highly beneficial. However, OSPF TE comes with its own drawbacks-the price one has to pay for higher through put is the increased
average delay caused by the significant increase in the regular path length.
Conclusion
We have shown that OSPF link weights produced by the Primal Minimum Cost Maximum Through put Problem supply
a reasonable basis to build a practical OSPF TE architecture onto. Such optimal link weights are certain to exist, integer valued, upper-bounded, and produce a route set that maximizes the throughput of the network. For logical sized networks the consequent ILP can be solved quickly. In addition, we introduced the notion of ideal distance to prioritize high-volume sessions and we future a scheme for tie flouting to pass up uncertainty in the shortest path illustration.
We consider that the framework obtainable in this letter is the first one that exhibits all the essential properties to make OSPF
TE a good choice for "poor man's traffic engineering."
