AN APPROACH TO SELF-MANAGEMENT BASED ON AUTOMATIC DIAGNOSTICS

2016 IEEE NetSoft Conference and Workshops (NetSoft), 2016

In this paper, we propose a multi-layer self-diagnosis framework for networking services within SDN and NFV environments. The framework encompasses three main contributions: 1) the definition of multi-layered templates to identify what to supervise while taking into account the physical, logical, virtual and service layers. These templates are also finergranular, extendable and machine-readable; 2) a self-modeling module that takes as input these templates, instantiates them and generates on-the-fly the diagnosis model that includes the physical, logical, and the virtual dependencies of networking services; 3) a service-aware root-cause analysis module that takes into account the networking services' views and their underlying network resources observations within the aforementioned layers. We also present extensive simulations to prove the fully automated, finer granularity and reduced uncertainty of the root cause of networking services failures and their underlying network resources.

2005

The communication and sharing of information is a fundamental need to computer users. The number of users who have needed to communicate with others has rapidly increased. Over the last decade, the network domains have become very complex structures. Hence, many faults may occur, which cause them to operate outside their required quality of service. The use of management techniques is required to handle the complexity of computer networks in order to detect and control the behaviour of system resources. Currently, the management systems run on a central station (client-server), where data needs to be transferred from the client to the server and vice versa. This may lead to inefficiency if the network is very large in scale. The use of Mobile Agents (MA) is a new technology based on mobile code that can migrate from one node to another in order to collect, analyse and make decisions locally or remotely on behalf of a centralized network management station.

Proceedings of INFOCOM '97, 2000

The increasing role of communication networks in today's society results in a demand for higher levels of network availability and reliability. At the same time, fault management is becoming more difficult due to the dynamic nature and heterogeneity of networks. We propose an intelligent monitoring system using adaptive statistical techniques. The system continually learns the normal behavior of the network and detects deviations from the norm. Within the monitoring system, the measurements are segmented, and features extracted from the segments are used to describe the normal behavior of the measurement variables. This information is combined in the structure of a Bayesian network. The proposed system is thereby able to detect unknown or unseen faults. Experimental results on real network data demonstrate that the proposed system can detect abnormal behavior before a fault actually occurs.

2004

The lack of specialized experts in diagnosing network faults, inconsistencies of diagnose results and professional opinions, time-consuming and growing complexity of this task; has motivated the dewlopment of our c\utomated Network Fault Diagnostic System. This system aims to serve as an intelligent diai-,'llOStic system that will be able to produce fast, accurate, user-friendly and appropriate suggestions that will assist normal network users and administrators respectively. To ensure the realistic and successful development of the system, we adopt Extreme Programming methodology. l\lany efforts have been paid to implement a novel and efficient solution to precisely diagnose problems and in timely manner. The methodology has e\·oh-ed from rule-based systems through case-based systems to more recent model-based systems. Our project is designed upon case-based diagnostic approach as it suggests the use of previously experienced, concrete problem or cases instead of rules or model...

Lecture Notes of the Institute for Computer Sciences, Social Informatics and Telecommunications Engineering, 2011

In this paper we discuss the perspectives that should be taken into account by the research community while trying to evolve Fault-Management towards Autonomic Fault-Management. The well known and established FCAPS Management Framework for Fault-management, Configuration-management, Accounting-management, Performance-management and Security-management, assumes the involvement of human technicians in the management of systems and networks as is the practice today. Due to the growing complexity of networks, services and the management of both, it is now widely believed within the academia and the industry that the concept of Self-Managing Networks will address some of the current challenges in the management of networks and services. Emerging Self-Management technologies are promising to reduce OPEX for the network operator. There is still a lot of work to be done before we can see advanced, production level self-manageability aspects of systems and networks, beyond what has been achieved through scripting based automation techniques that have been successfully applied to management and network operation processes. The concept of autonomicity-realized through control-loop structures and feedback mechanisms and processes, as well as the information/knowledge flow used to drive the control-loops), becomes an enabler for advanced self-manageability of networks and services, beyond what has been achieved through scripting based automation techniques. A control-loop can be introduced to bind the processes involved in each of the FCAPS areas, and the "autonomic manager components" that drive the control loops and are specific for different FCAPS should interwork with each other in order to close the gaps characterized by dependencies among FCAPS functional areas as the FCAPS functional areas go autonomic and realize self-management. The dependencies among FCAPS functional areas need to be studied such that the functions/operations and processes that belong to the different areas can be well interconnected to achieve global system goals, such as integrity, resilience and high degree guarantee of system and service availability.

SRI International, 1996

Traditionally, network management activities, such as fault management, have been performed with direct human involvement. However, these activities are becoming more demanding and data intensive, due to the heterogeneous nature and increasing size of networks today. ...

2001

In this paper, we present a framework for supporting intelligent fault and performance management for communication networks. Belief networks are taken as the basis for knowledge representation and inference under evidence. When using belief networks for diagnosis, we identify two questions: When can I say that I get the right diagnosis and stop? If right diagnosis has not been obtained yet, which test should I choose next? For the first question, we define the notion of right diagnosis via the introduction of intervention networks. For the second question, we formulate the decision making procedure using the framework of partially observable Markov decision processes. A heuristic dynamic strategy is proposed to solve this problem and the effectiveness is shown via simulation.

IEEE Internet Computing, 1998

s the Internet becomes a critical component of our society, a key challenge is to maintain network availability and reliability. The increasing number of networks is making downtime more costly. Network management tools and techniques must therefore become better at detecting and identifying network faults. Fault management is no easy task, however. Fundamental changes to the network occur much more frequently as demand grows and components and applications are developed in an open environment. Although mixing and matching hardware and software from different vendors supports customization and use of the latest technologies, it also increases the risk of faults and other problems. 1 Current fault management implementations generally rely on the expertise of a human network manager, which is translated to a set of rules and then to threshold levels on the measurement variables being collected. As networks become more complex and changes occur more frequently, the human network manager will find it hard to maintain a sufficient level of expertise on a particular network's behavior. Fault management research has covered approaches such as expert systems, finite state machines, advanced database techniques, and probabilistic methods. Aurel Lazar and colleagues give a good review of communication network fault detection and identification. 2 The drawback to all these approaches is that they require a specification of the faults to be detected, and it is not feasible to specify all possible faults. Also, changes in network configuration, applications, and traffic can alter the type and nature of possible faults, which makes modeling them impractical in many cases.

ACM SIGCOMM Computer Communication Review, 2009

By studying trouble tickets from small enterprise networks, we conclude that their operators need detailed fault diagnosis. That is, the diagnostic system should be able to diagnose not only generic faults (e.g., performance-related) but also application specific faults (e.g., error codes). It should also identify culprits at a fine granularity such as a process or firewall configuration. We build a system, called NetMedic, that enables detailed diagnosis by harnessing the rich information exposed by modern operating systems and applications. It formulates detailed diagnosis as an inference problem that more faithfully captures the behaviors and interactions of fine-grained network components such as processes. The primary challenge in solving this problem is inferring when a component might be impacting another. Our solution is based on an intuitive technique that uses the joint behavior of two components in the past to estimate the likelihood of them impacting one another in the present. We find that our deployed prototype is effective at diagnosing faults that we inject in a live environment. The faulty component is correctly identified as the most likely culprit in 80% of the cases and is almost always in the list of top five culprits.

1999

Abstract The lack of specialized professionals in network management and the growing complexity of this task has been aiming the need for developing tools to give support to the network administrator task. The construction of such tools requires an intense process of knowledge acquisition from experts in the area as well as the use of Artificial Intelligence (AI) techniques.