R. Gopal. Layered model for supporting fault isolation and recovery. In NOMS'00 [34].  Home/Search   Document Not in Database   Summary   Related Articles   Check

....entities. As events propagate up the entity model hierarchy, their semantics becomes more and more abstract so that higher level scenarios do not need to know the low level details of the network state to correlate low level faults with the symptoms observed in the higher layers. Gopal et al. [10] proposed a model for multi layer fault diagnosis. Although it was originally used to enhance the presentation of the system state on the management console for the purpose of manual fault localization, the model is also suitable for automatic fault localization with the use of model traversal ....

....on hosts a and b, and services that layer L 1 offers between hosts a and b. Layer L functions on node a depend on layer L 1 functions on node a. The recursive dependencies between services, protocols and functions constitute a dependency graph. Both Yemanja [2] and the layered system model [10] address the issue of vertical fault propagation. However, the fault propagation techniques proposed in [2, 10] are based on the assumption that, when higher level entity fails or experiences performance problems, it is easy to choose a small set of lower level entities which may be responsible ....

[Article contains additional citation context not shown here]

R. Gopal. Layered model for supporting fault isolation and recovery. In NOMS'00 [34].

.... that pertain to failure conditions on that level and generalize them into events with a higher degree of conceptuality using the concept of composite events [18, 19] This layered correlation model is in compliance with recent work on modeling networks for supporting fault isolation and recovery [8, 23]. We believe that the layered approach to scenario development has many advantages over other approaches, these include: Support for the development of reusable correlation scenarios. In most cases system reconfiguration does not necessitate scenario modification. Rule development is ....

....or layer failed may initiate automated recovery, notify the system administrator, and or open a problem record. Communication systems are frequently modeled using a dependency graph whose directed edges indicate that the entity at the tail of an edge depends on the entity at the head of the edge [8, 15]. The network impact graph created by reversing the edges of the dependency graph shows which entities may fail because of the failure of another entity [8] It may be observed that failure symptoms propagate along both dependency and impact graph edges. For example, layer two s ability to form a ....

[Article contains additional citation context not shown here]

R. Gopal. Layered model for supporting fault isolation and recovery. In J. W. Hong and R. Weihmayer, editors, Proceedings of IEEE/IFIP Network Operation and Management Symposium, Honolulu, Hawaii, Apr. 2000.

....6. A comparison of our solutions with other event correlation techniques is presented in Section 7. 2 Layered model for alarm correlation For the purpose of fault diagnosis, communication systems are frequently modeled in a layered fashion imitating the layered architecture of the modeled system [11, 35]. This approach provides a natural abstraction of the modeled system s entities, reusability of the model s modules, and ability to divide the fault management task into separate, simpler subtasks. Because of fault propagation, the effects of an abnormal operation of functions or services provided ....

....of fault propagation, the effects of an abnormal operation of functions or services provided by lower layers may be observed in higher layers. Fault management systems model fault propagation by representing either causal relationships among events [12, 35] or dependencies among system entities [11, 18]. In the layered fault model, the definition of entity dependencies is based on real life relationships between layers on a single host and among network nodes communicating within a single protocol layer. The model presented in this paper is based on the model proposed in [11] However, our ....

[Article contains additional citation context not shown here]

R. Gopal. Layered model for supporting fault isolation and recovery. In Proc. of Network Operation and Management Symposium, Honolulu, Hawaii, 2000.

....so that the network can continue to operate, which is the fast and automated way to restore interrupted connections. In general, it is implemented with protection switches that change position when the optical power gets below a certain threshold. Recent work on fault isolation can be found in [Gop00] network (re )configuration (so called restoration) that minimizes the impact of a fault by restoring the interrupted connections using spare equipment. This involves some processing to discover the best path to re route the connections. replacement of the failing component(s) Several ....

R. Gopal. Layered model for supporting fault isolation and recovery. IEEE/IFIP: Network Operations and Management Symposium NOMS, 2000.

.... services to end systems, a management system needs to efficiently and accurately identify the occurring network failures [13, 25] A common procedure is to correlate network or service layer symptoms; however, this process is usually impaired by the large number of a system s layers and parameters [1, 8, 22], their interactions, and the uncertainty about their state. This paper presents a preliminary study of applying statistical techniques [16, 18] known in the engineering quality control to cope with the exponential complexity that often hampers the event correlation process. The concept of ....

....(4) ability to learn event correlation patterns [14] and (5) temporal event correlation [15] Most of these techniques rely on the assumption that the existence of multiple simultaneous faults is negligible. Relationships between network objects are often represented using a dependency graph [8, 12, 13]. Dependency graph is a directed acyclic graph G(V, E) whose nodes V correspond to network objects (both physical and abstract) and edges E describe dependencies between the objects. Edge (v i , v j ) E represents the fact that object v i affects object v j , which we denote by v i v j . ....

R. Gopal. Layered model for supporting fault isolation and recovery. In Proc. IEEE/IFIP Network Operation and Management Symp., Honolulu, Hawaii, 2000.

....low level resource availability related problems such as a broken cable or an inactive interface. Recently, the scope of fault localization has been expanded to include the diagnosis of performance problems in higher layers of the communication system, such as the transport and application layers [1], 2] For this purpose, non deterministic reasoning needs to be incorporated in the fault localization process [3] This paper utilizes belief networks [4] to perform fault localization in communication systems whose failure propagation model may be described by bipartite causality or dependency ....

.... L on host H may also affect a system entity in layer L on host H # that communicates with host H (horizontal propagation) Fault management systems model fault propagation by representing either causal relationships among network events [7] 8] or dependencies among communication system entities [1], 6] 9] In the past, researchers focusing on fault localization frequently assumed a deterministic fault propagation model, in 0 7803 7476 2 02 17.00 (c) 2002 IEEE. which one can predict with certainty whether a failure of an entity will cause a failure of a dependent entity [7] 8] The ....

[Article contains additional citation context not shown here]

R. Gopal, "Layered model for supporting fault isolation and recovery," in Proc. of Network Operation and Management Symposium, Honolulu, Hawaii, 2000, pp. 729--742.

.... backup systems when a production server has to be brought down for performing a software upgrade) However, the main problem today lies in the fact that dependencies between services and applications are not made explicit, thus making root cause and impact analysis particularly difficult [3]. Solving this problem requires the determination and computation of dependencies between services and applications across different systems and domains, i.e. establishing a global service dependency model and enabling system administrators to navigate through the resulting directed graph from ....

R. Gopal. Layered Model for Supporting Fault Isolation and Recovery. In J.W. Hong and R. Weihmayer, editors, Proceedings of the 7th IEEE/IFIP Network Operations and Management Symposium (NOMS'

.... used in today s management world although their benefits are commonly known (section 2 explains existing applications) This fact leads to a lack of overview for the ITadministrators and prevents the use of powerful management tools like event correlators that are based on dependency models [2]. More applications are described in section 2 together with an overview of existing types of dependency models. Typical enterprise scenarios will of course be much larger than the example above. In such cases it would be hard to keep the overview even with the help of dependency models. To ....

R. Gopal, "Layered Model for Supporting Fault Isolation and Recovery, " In Hong and Weihmayer [21], pp. 729--742.

....a consequence, dependency models are not generally used in today s management world although their benefits are commonly known. This leads to a lack of overview for the ITadministrators and prevents the use of powerful management tools like event correlators that are based on dependency models [2]. More applications are described in section 2 together with an overview of existing types of dependency models. To overcome the problems of automated modeling described above, this paper presents a new approach to gain management relevant dependency information. Unlike conventional approaches it ....

R. Gopal, "Layered Model for Supporting Fault Isolation and Recovery," In Hong and Weihmayer

No context found.

R. Gopal. Layered Model for Supporting Fault Isolation and Recovery. In Proceedings of the 2000.

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC