- Number of study hours: 15
- Short description of the course: This module has two educative aims. The first is to describe the Change Management Process and to explain how it is located into the Service Support area. In particular, main concepts related to this process are the following: criticality of the IT change activities; definition of a base procedure for Change Management process, depending on ITIL® specification; description of connections among Change Management process and the other Service Support processes; definition and description of main elements of a Request For Change. The second one is describing methodologies and techniques employed in Network Management. The issue, undervalued at first, is gaining more and more importance in networking as a rapid individuation of bad operating disposals or preventing breakdowns through constant monitoring of meaning parameters, could lead to substantial performance improvement and saving of money. The module provides Network Management basic definitions and shows its fundamental components. Afterwards it presents protocol SNMP (Simple Network Management System) in detail, showing its fundamental components and its evolutions. At the end, the module presents some network management protocols alternative to SNMP, as for example RMON, and a few commercial solutions.
- Target groups: The employers of IT core level professionals are the target sector. The first target group consists of IT students (vocational school IT basic level training and the first courses of colleges and universities) in technology area and IT practitioners not having vocational certificates yet.
- Prerequisites: Prerequisite knowledge, necessary to the student in order to let him or her understand the module content, could be substantially listed as follows:
- Basic principles of Computer Network ways of working
- Knowledge about the Information Technology principles and terminology
- Knowledge of the “application level” in TCP/IP protocol architecture
- Knowledge of the “transport level” in TCP/IP protocol architecture, and of User Datagram Protocol (UDP) in particular
- Aim of the course - learning outcomes
- The expected knowledge are:
- Knowledge of the principal protocols in Network Management
- Knowledge of principal basic components involved in Network Management process
- Knowledge of principal Network Management techniques.
- The competence this module is expected to be furnishing rounds up in administration, maintenance and support of local and/or geographic computer networks. In particular, the outcoming professional (figures) is “High Network Supervisor/Technician”.
C.6.1 System ManagementEdit
C.6.1.1 Describe best practices in managing the configuration of an IT infrastructureEdit
The Configuration Management Process is responsible for the management and monitoring of the various configuration scheme applied to the IT infrastructure, during its lifecycle. This process manages the various components (CI ), which includes all the items that constitute the IT infrastructure (hardware, software, documentation, third-part provided services…). The main activities of this process are:
• Collection of information about each CI. • Analysis and definition of relations and dependences among the various CI. • Storing of collected information in a dedicated DB (CMDB). • System consistency checking, after each configuration change. • Continuous analysis and monitoring of IT infrastructure.
The Configuration Management Process provides a logical model of the IT infrastructure and services. It identifies, monitors, maintains and verifies the evolution of the various CI that form the infrastructure. The main aims of this process are the following:
• Provide support information to IT organization. • Define and document the procedures that must be used to manage CIs. • Provide an interface to the other management process, which can use it to obtain all CI information (stored in CMDB). • Assure that all CI changes are promptly reported. • Verify that the infrastructure includes only authorized CI. • Verify and guarantee the consistency of CMDB with respect to the authentic IT infrastructure configuration.
The CM process is often supported by another process, called Release Management Process, which provides a general view of the IT changes, considering all the aspects of a release (technical and juridical). Release Management is responsible for all the contractual and juridical commitments, related to the hardware and software components actually used into the IT infrastructure. To ensure that master copies of all software are secured, they are stored in a protected area called Definitive Software Library (DSL ). In the same manner, an area should be set aside for the secure storage of definitive hardware spares, the Definitive Hardware Store (DHS ).
C.6.1.2 Describe best practices in IT change and release managementEdit
Within an IT environment, the rapid evolution of technologies, that are the basis of services provided, and the business needs they support, involves need to change something in the IT infrastructure. Moreover, most of the actions taken to resolve problems or incidents involve some changes to components items (CI ) engaged in the failure. Every change applied in an IT infrastructure may represent a risk for the entire system. All these factors determine the need to plan a dedicated process, called Change Management process, which is responsible for managing changes in IT infrastructure. It is particularly important that Change Management processes have high visibility and open channels of communication in order to promote smooth transitions when changes take place.
In an IT context, change reasons can be various, for example:
• problem resolution • new IT services introduction • optimization of furnished services • cost reduction
Changes managed by Change Management Process can be related to all IT infrastructure components:
• hardware • communication instruments • net devices • software applications • software applications currently used (“live systems”) • support procedures and documentation
Note that developing applications are not part of change management field; they are managed by the Project Management Process. The person responsible for such process will take care to inform Configuration and Release Management Process managers on the management principles and new systems production release methods. In a business context supported by an IT service infrastructure, the causes that give rise to have a dedicated process for all the activities related to change management are:
• It is necessary to analyze and forecast the impact that IT changes can cause on business and, by turns, the impact that business changes can have to the IT infrastructure components. • It is need to individuate which problems cause the biggest changes. • Finally, natural business evolution and arising of new requirements can cause some changes in the IT infrastructure. These changes must be analyzed and scheduled when it’s possible. C.22.214.171.124 Change Management terminology, activities and CAB Terminology
According to ITIL ® definition, “change is the process of moving from one defined state to another”. Moreover, ITIL® adopts the following terminology to describe Change Management process:
• Change: a change is the addition, modification or removal of approved, supported or base lined hardware, network, software, application, environment, system, desktop build or associated documentation. • RFC (Request For Change): form, or screen, used to record details of a request for a Change to any Component Item within an infrastructure or to procedures and items associated with the infrastructure. • FSC (Forward Schedule of Changes): planning of changes that must be applied. It contains detailed information on all the approved Request For Changes with a precise change implementation scheduling.
The activities carried out by the process in order to manage the RFCs are the following:
• RFC logging and filtering: RFCs are analyzed, classified and stored. A priority is assigned to every RFC. • Planning: the management of RFCs is planned producing the FSCs. • RFC approval: the Change Advisory Board (CAB ) decides which changes can be applied. • Build and testing: the changes are applied and tested. • Post implementation review: Change Management must review all implemented Changes after a predefined period has elapsed. • RFC closure: at this time, changes can be considered definitive (Configuration and Release Management processes can update information on actual system status). • Management reporting.
The Change Advisory Board (CAB) is a body that exists to approve Changes and to assist Change Management in the assessment and prioritisation of changes. As and when a CAB is convened, its members should be chosen who are capable of ensuring that all Changes are adequately assessed from both a business and a technical viewpoint. To achieve this mix, the CAB needs to include people with a clear understanding of the Customer business needs and the User community, as well as the technical development and support functions. When major Problems arise, there may not be time to convene the full CAB, and it is therefore necessary to identify a smaller organisation with authority to make emergency decisions. Such a body is known as the CAB Emergency Committee (CAB/EC ).
C.126.96.36.199 Change Management process tasks
One of the main activities of the Change Management process is to log the applied changes and the change requests. Usually, all the RFCs managed are logged with the following information:
• RFC identification number • Actions and changes caused by RFC • Infrastructure components involved in changes • RFC status • RFC priority and classification
By classifying the RFCs, the request scheduling and planning and the activities carried out by the other Service Support process (especially for Configuration Management and Problem Management processes) are noticeably simplified.
In order to produce a good change scheduling, it is used a prioritization of the RFCs approved by CAB. Every RFC should be allocated a priority that is based on the impact of the Problem and the urgency for the remedy. The following priority ratings, provided as examples only, is suggested by ITIL®
• Immediate: causing loss of service or severe usability problems to a larger number of Users, a mission-critical system, or some equally serious problem. Immediate action required. • High: severely affecting some Users, or impacting upon a large number of Users. To be given highest priority for Change building, testing and implementation resources. • Medium: no severe impact, but rectification cannot be deferred until the next scheduled Release or upgrade. To be allocated medium priority for resources. • Low: a Change is justified and necessary, but can wait until the next scheduled Release or upgrade. To be allocated resources accordingly.
ITIL® suggests to assign a category to every RFC, selected on the potential change impact on system:
• Standard: a RFC is standard if the change activities are well-known and risk level is low. Related changes can be applied without CAB approval. • First level: it is assigned to low risk RFCs that are easily implementable. Change can be applied after approval by Change Manager. • Second level: this RCF have a medium potential risk and impact. They must be approved by CAB. • Third level: urgent changes with very high potential risk level and impact. CAB or CAB/EC must be urgently summoned for the changes approval.
C.6.1.3 Define the purpose of a request for change (RFC) and the essential elements that it should containEdit
RFC Contents and change impact analysis
The following items should be included in an RFC form: • RFC identification number • description and identity of item(s) to be changed • reason for Change • effect of not implementing the Change • version of item to be changed • name, location and telephone number of person proposing the Change • date that the Change was proposed • Change priority • CAB recommendations where appropriate (which may be held separately, with impact and resource assessments, where convenient) • scheduled implementation (Release identification and/or date and time) • back-out plan • risk assessment and management • impact on business continuity and contingency plans • status of RFC (i.e. ‘logged’, ‘assessed’, ‘rejected’, ‘accepted’, ‘sleeping’)
As before described, in order to estimate the emergency and priority of a RFC, it is important to evaluate impact that a change cans have both on business, and on IT infrastructure. This impact depends on several factors, which must be valuated and analyzed: • the effect upon the infrastructure and Customer service, as defined in the SLA • the impact on other services that run on the same infrastructure • the impact on non-IT infrastructures within the organisation – for example, security, office services, transport, business – Customer Help Desks • the effect of not implementing the Change • the IT, business and other resources required to implement the Change • additional ongoing resources required if the Change is implemented. Aim of this analysis is to determine if the benefits produced by a change justify the related risks. Depending on the results of this analysis, the CAB decides the approval of RFCs.
C.188.8.131.52 Process benefits and change implementation
Specific benefits of an effective Change Management system include:
• better alignment of IT services to business requirements • increased visibility and communication of Changes to both business and service-support staff • improved risk assessment • a reduced adverse impact of Changes on the quality of services and on SLAs • better assessment of the cost of proposed Changes before they are incurred • improved Problem and Availability Management through the use of valuable management information relating to changes accumulated through the Change Management process • increased productivity of IT staff through less need for diversion from planned duties to implement urgent Changes or back-out erroneous Changes • an enhanced business perception of IT through an improved quality of service and a professional approach.
Although it may be better to implement one Change at a time this is not usually practical. Wherever possible, Change Management should schedule approved Changes into target Releases and recommend the allocation of resources accordingly. There is clear continuity between the Change Management and Release Management processes. Release Management processes impact upon the Change management process and, in particular, have a role in developing and maintaining standard Changes that introduce new or revised software and hardware into the infrastructure. As Releases are the manifestation of Changes, the Change process initiates the Releases under the agreed, documented and maintained Release process.
Authorised RFCs should be passed to the relevant technical groups for building of the Changes. Change management has a coordination role to ensure that these activities are both resourced and also completed to schedule. It is important to ensure that the same standards and methods that were used for an original component are again used for the Change. Back-out procedures should be prepared and documented in advance, for each authorised Change, so that if errors occur after implementation, these procedures can be quickly activated with minimum impact on service quality. To prevent Changes from adversely impacting on service quality, it is strongly recommended that Changes be thoroughly tested in advance (including back-out procedures where possible).
C.184.108.40.206 Process evaluation Change Management process is strictly related to Configuration Management process and the other processes in both Service Support and Service Delivery areas (especially Release Management and Capacity Management). Every change applied by Change Management process must be notified to Configuration Management process, in order to update CI information contained in CMDB . In this way, it is also possible to detect any unauthorized change. Moreover, the information stored in CMDB is vital to risk and impact assessment activities. By analysis of information about CI involved in a RFC, the Change Management process is able to estimate both the risk related to a change, and the other CI eventually involved in the change. In this way is possible to know if a RFC satisfaction (that is the application of the change) can raises further changes and, therefore, further RFC.
In order to evaluate Change Management performances, there are periodically created some management reports containing all the activities performed by the process in the period. Consider including the following facts and statistics in management reports:
• Number of Changes implemented in the period • A breakdown of the reasons for Change • Number of Changes successful • Number of Changes backed-out, together with the reasons • Number of Changes caused by incidents • Total number of RFCs • Number of RFCs rejected • Number of CIs involved in changes • High incidences of RFCs relating to one CI
The information can be used as a basis for assessing the efficiency and effectiveness of the Change Management process. For this, it is necessary to filter out effects that are outside the direct control of Change Management. For example, frequent Changes affecting a particular CI may be a result of fragility of that item, and should not reflect badly on Change Management as might at first be inferred. Similarly, frequent Changes in User facilities may reflect a rapidly changing User requirement.
C.6.2 Principles of Network ManagementEdit
C.6.2.1 Main functions of a network management systemEdit
As the hardware and software components constituting the network have been installed and integrated correctly, they need a continuous administrative and maintenance process in order to preserve an adequate functionality and the required performance level. Besides, network devices need a continual monitoring in order to grant the execution of administrational politics concerning network usage and security. When networks were introduced for the first time the issue of their management did not manifest its critical urgency fully. Then, they were considered a mere research field without being the wide spread used tool as it is nowadays. Together with the extensive diffusion and transformation of computer networks in local/private networks – the so-called “intranet” – they evolved from networks providing a few workplaces to essential infrastructures whose working is of the utmost importance in the ordinary carrying out of production activities. As a consequence, a more systematic administration of the large number of hardware and software components involved in the process has become of great importance. C.220.127.116.11 Description of the main components and functions of a Network Management System The following statement can be considered a sufficiently general definition for Network Management: “Network Management consists in the operating, integration and coordination of human, hardware and software elements as to supervising, examining, configuring, analysing, evaluating and controlling the network in order to satisfy – at a reasonable cost – operational performances and quality requirements necessary for the services the network intend to supply”. This definition understands to the reference architecture in Figure 1. The principal elements in this architecture are:
• Network Management Console • Managed Devices and its related “agent” • Management Proxy
Moreover, as shown in Figure 1, different devices refers all to a standard modality for monitoring data storage by means of a Management Information Base (MIB ).
Figure 1: General Architecture in a Network Management System The principal elements involved in a network management tool are:
• Network Management Console • Managed Devices and its related “agent”
The Network Management Console is one of the most important element in a network management architecture. The Console is usually constituted by a Workstation supplying the necessary interfaces for network management administration and hosting a management entity, which forms indeed the management system core. The Console allows every network manager to constantly monitor the network status, and checks the devices proper working as well as their correct configuration. Among some typical features available on the Console interfaces, there is the possibility to check the network supported traffic load and the available resources. Figure 2 shows a management interface exemplary screenshot.
Figure 2: Network Management Software Interface Screenshot.
Among the Managed Devices there are all the active network devices such as workstations, servers and other network devices as hubs, switches and routers. These devices host an appropriate management software module (called agent) which gathers performance and working information. Agents memorize these information inside a MIB (Management Information Base). Thus, whenever requested, agents can send the management entities the information gathered and stored within the network management console. The block of agents, MIB and Console are often named Network Management System (NMS) A MIB is nothing less than a properly defined database into whose table merge all the information gathered by the agents. MIB does not hold any application logic but just a definition of the information the agent is asked to be gathering, together with the relative setting that make the exchange of information with other devices possible. The displacement of this sort of data from MIB to NMS control center takes place by a network management protocol as, for example, SNMP protocol (Simple Network Management Protocol). Whenever the agents determine a malfunctioning in the device they are monitoring, they instantly send an alarm to the management entity. The entity, as soon as it receives the alarm, can activate one or more events as, for example, send the operator a notification, record the event, close the system, try to isolate and fix the failure automatically. The same management entity can also decide to question the agents placed in the terminal station so to acquire further information on determined variable. This questioning can be automatic or explicitly requested by an operator.
C.6.2.2 Different Parameters which can be managed in a networkEdit
C.18.104.22.168 Performance Managing ISO (International Standard Organization) formalized all managing functions in five general categories: • Performance Managing • Configuration Managing • Accounting Managing • Breakdown Managing • Security Managing Performance Managing category aims quantifying, measuring, report making, analyzing, and controlling different network devices performances. In practice, it consists in monitoring parameters as network productivity, responding time to user, and so on. Hence, this category is essential to collecting traffic statistics and, as a consequence, to applying methods against congestion and for service quality monitoring.
Figure 3 Network Managing Software Interface
C.22.214.171.124 Configuration Managing “Configuration Managing” general objective is monitoring network configuration data and devices, as to control and manage malfunctioning effects on hardware and software element composing it. It is important that network managing activity is joined to a configuration managing steady process, enriched by the possibility to control network status by reading and setting out devices configuration parameters constituting the network itself. Parameters that can be managed are, for instance, user bandwidth at connection setting, any specific functions assignment under user request, and so on. Configuration managing also employs a database for housing all acquired data, which can be interviewed at any time:
Figure 4 Graphic Interface for a network configuration
C.126.96.36.199 Accounting Managing Accounting Managing aims to quantifying network parameters use for properly charging network usage to any single user or group of users. Thus, accounting managing deals with network resource usage statistics measuring, collecting, and recording. Accounting managing also allows network administrator to specify records and to monitor users and network devices access to resources. C.188.8.131.52 Breakdown Managing An informative “black-out” due to a LAN malfunctioning could be very expensive for a company in terms of working hours lost as well as in terms of image damaging. Failures managing, thus, becomes one of the most important and implemented network managing functions. Failures managing essentially provides the following operations: • Breakdowns survey, or degrading symptoms survey by means of alarm surveillance; • Automatic – or through the agency of the administrator – determination of breakdowns origin and cause. The use of breakdowns isolation techniques is crucial in this phase; • Possible solution to breakdown determination. As the failure is corrected, it would be appropriate memorizing the breakdown typology and the given solution within a specific database; • Correct network functioning control as soon as breakdowns correcting operations have ended.
Figure 5 Network breakdown managing and fixing strategies pattern C.184.108.40.206 Security Managing Absolute network administrator priority is given to managing mechanisms able to grant network security. The increasing necessity for letting users and remote staff enter the business LAN by external workstations in fact raises security problems entity. Security managing aims to establishing and respecting control indication for network resources access by means of:
• Security principles and politics development; • Techniques – as, for instance, cryptography – for specific traffic type managing application; • Access authentication procedures activation; • Protection systems – such as firewall or IDS – by illegal access introduction and configuration; • Anti-Virus services and applications introduction.
C.6.2.3 Other system architectures for network managementEdit
The architecture of a Network Management system described in Figure 1 can be considered universal. Nevertheless, its entire implementation can result too much expensive or complex to achieve.
Different and simplified approaches are adoptable when the network to monitor and manage is relatively small or the available resources are scarce. In such situations, a possible solution is to supply the network manager simply and cheap software tools for “sniffing” the network traffic. Such tools can be made available on a dedicated workstation or can be directly installed on some of the monitored network devices. The network manager can use the quoted tools to collect statistics about the network traffic so achieving an “indirect” monitoring capability. Starting from this “rough” monitoring, he can also decide and implement a control strategy to be realized through specific interventions on the network nodes.
Even if the described approach can appear very poor, its effectiveness and its cost/benefit ratio can be very good for many real world scenarios. Figure 6 depicts a possible example of use of the quoted architecture.
Figure 6 An example of LAN monitored by the use of a Sniffer Tool
C.6.3 The Simple Network Management ProtocolEdit
C.6.3.1 Main Components of the Simple Network Management ProtocolEdit
As it has already been stated, networks and processing systems have become predominant in firms, organizations and, more generally, in everyday life. As a consequence, a set of services and instruments for a more or less automatic network managing have been implemented. This unit introduces a network managing integrated system with a single interface for network managing by a set of commands. In particular, the key elements in a network managing system are:
• A managing station • An agent • The Informative base for managing • A network managing protocol
The first three elements have already been presented, it is now time for focusing on the network managing protocol. The managing station and the agents communicate by an apposite network managing protocol. In TCP /IP network managing it is named SNMP (Simple Network Management Protocol).
C.220.127.116.11 Basic Characteristics for a Network Managing Protocol
A network managing protocol comprises the following key functions: drawing, setting out and notifying/advising. The drawing function consents the managing station to recover the data value included in MIB . The setting out function consents the agent to communicate to managing station remarkable events occurred on the monitored component. There are different communicating protocols for network devices data collecting. In 1988 SNMP specific was published and it has gradually become the predominant standard for network managing. SNMP protocol was originally conceived for network devices interconnections in TCP/IP. Today, thanks to the various modifications and revision it has gone through, SNMP applies to any network.
C.6.3.2 Introduction to SNMP ProtocolEdit
SNMP protocol, as it is clear by its name, aims to be a simple instrument for network managing. In particular, it defines a limited MIB, easy achievable and manageable, and an efficient protocol for recovering and setting MIB variables by the network manager. It is also provided with a device, called trap, by which agents can send notifications even without the explicit request by the managing station. Simplicity is SMNP strength as it consents easy implementation with reduced processing ability and scarce network resources. The simplicity of this protocol structure is not to be underestimated as it consents an easy interconnection among managing stations, agents, and different network components manufacturers. The main SNMP imperfections are related to security issues. The SNMP protocol was originally conceived for interconnecting devices in TCP/IP environment but after several revisions it now applies to any network environment. It implies that every network device houses data agent whose task is collecting data and sending them to the managing console as messages. Protocol SNMP manages this data exchange and defines 5 different kind of messages for accessing to managing data. This data exchange occurs in a Client-Server relationship where the network managing station represents the client and the monitoring agent represents the server. Each message is defined in separate PDU (Protocol data Unit) and arranged as it follows:
• GetRequest • GetNextRequest • SetRequest • GetResponse • Trap
Figure 7: PDU Example All previously listed messages can be described as follows:
• GetRequest: this message type is the most used and it is usually sent to the agent by the network manager. Its task is identifying the value of a specific managing variable. • GetNextRequest: this message is directed from the managing station to the agent and aims to obtain the value of the items included in a managing data table whenever the network manager does not know the variables in the MIB at the occurrence of a specific event • Response: this message is sent to the managing station by the agent further to a previous request • Set Request: this message is sent to the agents by the station in order to create, memorie, or modify a variable. The agent responds necessarily to each message with a Response message. • Trap: Trap is sent by the agent to the managing station so that the station can withdraw an event whenever necessary. This operation occurs with no request by the managing station. Some Trap messages have previously defined significance: • Cold Start: the managed device undergoes a total reboot which could lead to its reconfiguration • Warm Start: the device runs a reboot “on the spot” which does not change the existing configuration • Link Down: one of the links connected to the device is not operative • Link Up: a previous out of order connection has been restored • Authentication Failure: a Workstation is trying to communicate with a device but could not correctly authenticate • EGP Neighbor Loss: communication with a device connected by means of the EGP (Exterior Gateway Protocol) is failed
C.6.3.3 Main Limitations of the SNMP protocolEdit
SNMP is conceived as an application level protocol in TCP/IP protocol suite, and its action is based on transport level protocol UDP . Such an implementation choice implies that SNMP is a connectionless protocol and that each message exchange between agent and control station is to be considered a different transaction. As it has already been stated, while SMNP has its strength in its simplicity and well-defined structure, on the other hand, it also exhibits unsatisfying performances. One of the crucial problems is network traffic generated by messages exchange among the agents and the central network managing station, with a remarkable overload of the station. The problem derives from centralized network managing suggested by protocol SNMP. In order to solve the problem, evolutions of the protocol, named SNMPv2 e SNMPv3 have been introduced.
Figure 8: SNMP Architecture C18.104.22.168 SNMPv2
Protocol SNMPv2, against all expectations, does not manage the network directly but creates a framework for developing the applications for network managing. In other words, SNMPv2 furnishes the infrastructure for network managing. This version of SNMP , is involved essentially in routing data housed in MIB among the managing entity and the agents attending on request. SNMPv2 main part is still housed in the protocol used for exchanging network managing data. In practice, every participant in the network managing system has its local data base, MIB, where data are stored, and SNMPv2 defines a standard structure known as Structure of Management Information (SMI ). General architecture of SNMPv2 provides for at least a system responsible for managing, and housing managing applications. As a matter of fact, more than one station of that kind is needed in account of two fundamental aspects in extended network managing: redundancy and balanced task distribution. Other systems in the network run as collecting data agents that memorize data in MIB so that the managing responsible systems could then analyze them. The stored information contains data referred to the system itself or to the network, or networks, traffic to which the agent is connected. SNMPv2 supports a centralized network managing together with a distributed managing. There are systems, actually, that can operate as agents as well as administrators. In practise, an agent could respond to a managing system whenever the latter asks for local MIB data while it could also act as a proxy when requests are about remote devices.
Figure 9: Network managing through protocol SNMPv2 use Managing systems can question the agents for acquiring data from that system MIB, or for asking data about remote systems. In that case, the agent runs as a proxy undertaking the role of administrator for accessing data held by a remote agent, and then becomes again an agent sending data to the manager that has made request for. These exchanges are coordinated by the work of protocol SNMPv2. The protocol is based on a question/answer approach and makes use, as SNMP does, of transport level protocol, of suite TCP /IP , and of UDP . Data exchanges referred to SNMPv2 are in form of double question/answer and they do not need reliable connections, which are not supported by UDP. As already said, messages are usually employed for finding, modifying and setting up values associated with the MIB of the system to be managed. The protocol also affords trap messages which allow an agent to send unsolicited messages whose task is notifying an exceptional event – as, for instance, a link congestion, an out-of-order interface, or an apparatus congestion – to a managing entity. As previously stated, the most peculiar aspect in SNMPv2 lies in the protocol capacity to afford a clear and well defined mechanism for data exchange among all the elements constituting the network managing system. The system components communicate by messages constituted by an external heading and by a Protocol Data Unit (PDU). The external heading has some security implications which will be analyzed afterwards. For what concerns PDU, it is worth saying that SNMPv2 defines seven PDU typologies, formally defined by ASN.1 language.
Figure 10: Protocol SNMPv2 cases of PDU SNMPv2 defines seven different messages – similar to those considered for SNMP – that will be described later. GetRequest type PDU is generated by a manager and includes one or more object names whose value has been requested. The request is forwarded to an agent able to respond by a Response type PDU. In such a case, the field variable-binding contains the list of IDs and the values of all the requested objects. In case the requested object is not present, the Response PDU will contain the ID and a specific fault code. In this way, the agent can send an answer to the manager, though incomplete, so introducing an important improvement in respect to classical SNMP. As a matter of fact, in the previous protocol version whenever one or more variables in a GetRequest are not supported, the agent answered with a message containing a laconic noSuchName. Another message sent by the manager to controlling agents is GetNextRequest including a list of one or more objects. For every object mentioned in the field variable-binding, the next value in lexicographic order is returned. In other words, the object subsequent to that indicated in terms of position within the object ID tree-structure is returned. Obviously, the agent will try to return as much requested variables as possible. Another important characteristic of this kind of PDU lies in the fact that it allows a managing entity to dynamically find a MIB structure. This propriety could be determining in the case of a manager not knowing how many and which objects a specific agent is controlling or are present in a particular MIB representation. One of the most important extension introduced by protocol SNMPv2 in comparison with SNMP is GetBulkRequest PDU. This message type minimizes the number of protocol data exchange in case the manager and the agents should exchange a big quantity of data. GetBulkRequest enables a significant group of data to be returned avoiding the necessary redundancies due to consequential repetition of messages of the type GetRequest and GetNextRequest. SetRequest is another message delivered by the manager and directed to the agents. Its aim is asking for a modification of one or more objects. The agent receiving this request answers with a Response message containing the same Request-ID. SetRequest working principle is quite simple as only two cases are possible. In case of success all variables are updated and Response PDU contains variable-binding fields with a value for each variable. In case even one value cannot be set for whatever reason, the Response PDU will contain the reason for failure in error status index and the variable causing the failure in error index range. A very important message for the correct functioning of protocol SNMPv2 is the message named SNMPv2-Trap. It is generated and transmitted whenever an anomaly occurs in the network and it enables the managing station to receive asynchronous messages by the controlling agents. Unlike most protocol type messages, TRAP message do not need any answer by the receiving unit. In practice, that type of message is not confirmed. As for protocol SNMP, there are different TRAP messages able to model malfunctioning causes. InformRequest PDU is sent by a SNMPv2 entity itself operating in quality of manager in order to furnish managing data to the application using the previous entity. The manager which receives a InformRequest message must confirm the reception with a Response PDU. Both in case of SNMP-Trap as well as of Inform-Request messages it is possible to define conditions for generating a notification and sending data. C.22.214.171.124 SNMPv3
Many designing and functional lacks in SNMP protocol have been solved with the introduction of SNMPv2. But this protocol was not able to overcome the limits regarding security and managing functionality. In January 1998 standard SNMPv3 was published with the clear intent to solve the problem. Introducing the new standard was meant to give an answer about security lacks, which was the reason why protocol SNMP was more used as a monitoring device than as control device (in practise, message SetRequest was rarely used). SNMPv3 provides three key services: authentication, confidentiality, and access control. The first two services are associated with USM (User-based Security Model), while the last service is defined in View Based Access Model (VACM ). Before describing in details protocol SNMPv3, it is worth focusing on its structural definition, analysing the information provided by document RFC 2571. SNMP applications are composed essentially by the following components: a command receiver, a notification receiver, a proxy forwarder, a command generator, a notification generator. A managing entity generally generates messages of the type GetRequest, GetNextRequest, GetBulkRequest, SetRequest PDU . PDU generated by SNMP application passes through the so called SNMPv3 engine, which is essentially composed by a dispatcher, a message processing subsystem, a security subsystem and an access control subsystem.
Figure 11: SNMPv3 engine and its applications A PDU generated by the application passes into SNMP engine and in particular through block shipment, whose task is determining SNMP version. Downstream this block lies the Message Processing System, where a message heading containing SNMP version number, a ID message, and message dimensional data are assigned to PDU. In case coding or authentication techniques are involved, SNMP message (PDU plus the previously introduced headings) passes through a further block named by security, where proper heading fields are added. The resulting message is passed to the appropriate transport protocol – usually UDP – while the default port number for SNMP is 161. Port 162 is normally used for TRAP message transmission. As already stated in previous units, SNMP together with monitoring network status is also able to control it. This functionality obviously exposes the network to external attacks. An intruder could be able to intercept a SNMP message or generate new ones on its own damaging the entire network. The solution to this problem is sending SNMP messages by proper security systems. SNMPv3 provides three security services: authentication, confidentiality – these two items belong to USM – and access control – defined in View Based Access Model. The authentication mechanism defined in USM aims at verifying that the received message was effectively sent by the entity whose ID is present in the message headline. This inspection guarantees that the message was not modified, delayed or repeated during the transit. In order to get to the objective, SNMP combines the use of a HASH function with a secret key value in accordance with HMAC (Hashed Message Authentication Code) approach. The authentication mechanism is quite simple: the sender includes an activation code in the sending SMNP message. That code is functional to message content, to terminal sending system ID, to terminal receiving system ID, to transmission instant, and to recipient. Network (or the structure the network is included) manager is responsible for keys distribution. These keys will be then memorized in the managers and agents databases. When the recipient receives the message, the same secret key is used in order to recalculate the authentication code. If the codes correspond, no alteration occurred; if the codes differ message could have been manipulated. USM discretion service enables the agents to cipher their communication by means of an apposite secret key. For this operation algorithm DES (Data Encryption Standard) is normally employed. Access control service sets up the agents so that they can offer a different access level to MIB managing different access policy. There are substantially two ways for limiting access to MIB. First, access could be limited only to part of MIB. For instance, a politic could be enabling all the managers to access data such as performance statistics, and enabling only a part of them to modify or read their configuration parameters. A second approach consists in limiting the operation a manager can perform on a part of MIB.
C.126.96.36.199 More Network Managing Protocols – RMON and SMON
Although SNMP is a simple and effective network managing protocol – especially thanks to all the revisions occurred – collecting data procedures increase network traffic and could saturate central managing server. In consideration of that, IETF developed an alternative managing pattern named RMON. MIB RMON basic specifics enable the implementation of a distributed managing system constituted by a RMON manager and a RMON probe. RMON probe is responsible for managing data collecting and it could be a real hardware device or a software application built-in the managing device. RMON manager, on the other hand, collects and elaborates data for presenting them to the system administrator. RMON specifics leave network administrators free to select monitoring probes more suitable to network control. Managing applications do not communicate directly with the managed device but pass through the probe RMON agent by protocol SNMP. Specifics of RMON MIB define a common format for data so that the manager and the probe could exchange data. It is worth noting that RMON does not define how the probe should collect data or how the manager should format data for presentation. RMON defines data in nine groups of monitoring elements, and each group provides specific data blocks in order to have the necessary monitoring requisites. A manager and a RMON agent are compatible when they can implement a combination of those nine groups (not necessarily all of them).
RMON Group Function Statistics It holds statistic data measured by the probe for each interface History It records and memorizes periodical network samples Alarm It compares periodically statistic samples with previously configurated spans/thresholds. In case of these parameters overtaking, an event is generated Host It contains statistic data associated with each Host detected in the network HostTopN This Host group extension lists the N devices which generates more statistics Matrix It memorize statistics for messages exchange among couple of addresses. Filters It consents RMON probe searching for corresponding packets to a determined filter defined by user Racket Capture It consents packets capture after their passing in a canal Events It controls events generation and notification from the device Groups/Functions
In 1999 IETF defined MIB SMON in document RFC 2613. The document intent was providing monitoring functionality for local commuting networks. When RMON was introduced local networks were composed by short network segments interconnected by bridges or routers. Thus it was possible monitoring a network just placing some probes on the principal segments. As there was no connection among each single segment, RMON could operate independently on each interface. When switches were introduced, segments number increased notably; thus a network administrator could see only a short part of the network and was enabled to monitor few segments at a time. This problem degenerated with the introduction of virtual local networks which do not permit to a traditional RMON probe to visualize the whole traffic in the local network. So, SMON is a RMON extension which enables visualizing by remote the traffic passing through a switch. The main contrbution of SMON to standard RMON is the possibility to define physical entities – as, for instance, a switch – and logical entities – as virtual network – as it were data source valid also for other RMON groups. In practice, SMON considers a switch as a monitored entity instead of considering it a group of local network segment. MIB SMON provides an introductive data structure which enlists available data origins on the switch and their functionalities.
C.6.4 Tools for System and Network ManagementEdit
C.6.4.1 Tools for Network ManagingEdit
It seems clear that the activity of managing performances and functioning of a network needs Tools able to rapidly and simply monitor and configure the network components. It is therefore understandable how fundamental is the introduction of hardware and software Tools able to put in practice whatever protocols promise. These Tools enable network managers to observe network performances in real time, localizing problems and providing a detailed description of such problems. The two main diagnostic Tools for networks are the physical level analyzers and protocol analyzers – also known as network analyzers. Physical level analyzers control transmission line integrity, while protocol analyzers examine frame, packets, and connection data transfer integrity. Some of these devices record parameters, as for instance transmission rate, error rate, and eventual transmission delay.
C.6.4.2 Most used software toolsEdit
Among the most popular systems for heterogeneous network managing there are the systems introduced by Hewlett-Packard, by Aprisa and by Sun, but many are the network software and hardware producers which contemplate similar application in their catalogues. Furthermore, various open source implementations are available, for example: Openxtra, Net-SNMP (an open source SNMP implementation), Netsnmpj (an open source SNMP for Java). In general, there software products adopt a distributed architecture where managing servers acquire information about the managing domain state of health. A centralized managing system collects data by these servers, analyses them, and decides control acts. Obviously, all these operations accomplish by the use of protocol SNMP, and these systems provide automatic assistance services for any correlation between event and implemented solution. HP OpenView Network Node Manager (NNM) is an applicative for network managing providing simplified and thorough connections vistas/perspectives by means of intuitive graphic interfaces. In this way network administrators can easily identify cases where their intervention is necessary. Functionalities provided by NNM application could be summed up as follows:
• Local and wide area network TCP /IP, IPX , and level 2 devices automatic location • Automatic events correlation as to point out network problem causes and enable administrator to examining all the events leading to a specific alarm • Preventive managing by analysis of previously acquired information • Tolerance to failures functionality which enables to program critical data backup for network managing
C.6.4.3 Managing SystemsEdit
Systems Management should provide the monitoring, diagnostic and automated error recovery to enable fast detection and resolution of potential and actual IT failure. Systems Management tools are an essential building block for IT Services that require a high level of Availability and can provide an invaluable role in reducing the amount of downtime incurred. The provision of Systems Management tools positively influences the levels of Availability that can be delivered. Implementation and exploitation should have strong focus on achieving high Availability and enhanced recovery objectives.
Systems management is strongly influenced by network management; it may involve one or more of the following tasks:
• Hardware inventories. • Server availability monitoring and metrics. • Software inventory and installation. • Anti-virus and anti-malware management. • User's activities monitoring. • Capacity monitoring. • Security management. • Storage management. • Network capacity and utilization monitoring.
So, we can assume that Network Management is part of System Management area.
C.6.4.4 System Management goalEdit
A system management utility has the following three key objectives: • Provide policy-based system management and service administration • Provision services to customers on demand and according to service level agreements (SLAs) with the customers • Leverage economies of scale
Policy-based system management means adjusting the service behavior according to some customizable policy. The policy may be customer-, user-, or machine-specific. It can also be a policy specified by the utility. An example of a policy is performing incremental backup on a daily basis and a full backup on a monthly basis. At the machine or user level, the policy might specify the exact time of the day when the backup is to be performed. "On demand" service provisioning means being able to configure and deliver a service just in time when the need arises. "SLA-driven" means that the quality of service is maintained at a certain prescribed level. An example of an on demand service is performing a non-routine backup just before upgrading the operating system of a machine. An example of an SLA-driven service is maintaining a particular machine configuration in a usable state 99 percent of the time during normal working hours over a specified period of time. To satisfy the first two objectives just described, for each kind of system management service, a utility has to manage multiple types of system management services (i.e., similar service functionality offered by different brands, versions, etc.) and needs to be capable of configuring and deploying different variants of a particular service, depending on the situation at hand. The third objective is driven mostly by pragmatic considerations. Leveraging economies of scale means being able to amortize costs efficiently over a large customer base. In other words, costs associated with system management services should grow sublinearly as a function of the number of customers, machines, and users.
Clearly, for a utility to be successful as a business concept, large-scale automation is required to achieve the three key objectives listed above. Large-scale automation in configuring and coordinating services is necessary to provide system management services from a utility.
C.6.4.5 State-of-the-art in system management technologyEdit
The companies that are the most established and largest in the systems management industry are IBM Tivoli, CA Inc., HP, and BMC Software. The Microsoft Systems Management Server (SMS) and the Tivoli Configuration Manager (TCM) are two widely used applications for managing systems. These applications provide mechanisms that facilitate the performing of certain management tasks by an IT administrator. Management applications typically provide mechanisms for gathering hardware and software inventory and performing software installation and distribution, license management, troubleshooting, and so on. These tools enable an IT administrator to perform management tasks from a central console. However, the management applications do not alleviate the need for an IT administrator to spend time thinking and developing steps needed to accomplish the task at hand. For instance, if an IT administrator needs to deploy a new version of software, then he or she has to develop an action plan that takes into account hardware and software prerequisites, best practices, and old versions of the software that are deployed. The seemingly simple task of deploying software translates into several interdependent steps that may vary, depending upon the configuration of each managed object.
C.6.4.6 Evaluate the system requirements for operating a network management toolEdit
In general a network management solution provides a set of monitoring and configuration tools for administering net components like switches, routers, hubs, hosts and access servers. A complete management solution should be expected to: • easily download files to a central staging directory • silently and automatically deploy files and updates to remote and local users • provide a management console for viewing and changing settings throughout the domain • work in a multi-platform environment • provide a locking mechanism to protect configurations So the following network management applications have to be provided: • Applications supplying graphical back and front panel views of the net devices. The dynamic and color-coded graphical displays simplify device-status monitoring, device-specific component diagnostics, and device configuration; • Network management software that allows network discovery, mapping, monitoring, and alarm tracking; • A suite of web-based network management tools that enables administrators to collect the monitoring and fault information needed to track devices critical to the network. From the hardware point of view, the system requirements for this scenario are the following: • Server System: a workstation with an adequate quantity of memory (both central and secondary). The server system has to manage connections with monitored devices and to apply the corrective actions. • DB Server System: it has to collect data coming from monitored devices. So a huge quantity of hard disk storage is needed. • Client system: normal personal computers and more in general devices that will be controlled. A multiplaform approach has to be supported. Client systems should have a bufferization mechanism in order to collect information even when network connection with server is down. 7. Links to additional materials: [BOU97] Boutaba, R., El Guemhioui, K., Dini, P., “An Outlook on Intranet Management”, IEEE Communication Magazine, vol. 35, pp. 92-99, 1997.
[CAS96] Case, T.L., Smith, L.D., Managing Local Area Network, McGraw-Hill, 1996
[ITU97] ITU-T Recommendation X.701, Information Technology-Open Systems Interconnection-Systems Management Overview, 1997 [ITU92] ITU-T Recommendation X.733, Systems Management-Alarm Reporting Functions, 1992 [ITU92bis] ITU-T Recommendation X.734, Systems Management-Event Report Management Functions, 1992 [KAI03] Kaiser, G., Local Area Networks, McGraw-Hill, 2003. [RAM98] Raman, L., “OSI Systems and Networks Management”, IEEE Communication Magazine, Vol. 36, pp. 46-53, 1998. [RFC91] RFC 1271, S. Waldbusser, “Remote Network Monitoring Management Information Base”, Nov. 1991. [RFC95] RFC 1757, S. Waldbusser, “Remote Network Monitoring Management Information Base”, Feb. 1995. [RFC96] RFC 1905, J.Case, K. McCloghrie, M. Rose, S. Waldbusser, “Protocol Operation for Version 2 of the Simple Network Management Protocol (SNMPv2)”, Jan. 1996. [RFC96bis] RFC 1906 J.Case, K. McCloghrie, M. Rose, S. Waldbusser, “Transport Mappings for Version 2 of of the Simple Network Management Protocol (SNMPv2)”, Jan 1996.
[RFC96ter] RFC 1907, J.Case, K. McCloghrie, M. Rose, S. Waldbusser, “Management Information Base for Version 2 of the Simple Network Management Protocol (SNMPv2)”, Jan. 1996
[RFC99] RFC 2570, J. Case, r. Mundy, D. Partain, B. Stewart, “Introduction to Version 3 of the Internet-standard Network Management Framework” May. 1999. [RFC99bis] RFC 2571, B. Wijnen, D. Harrington, R. Presuhn, “An Architecture for Describing SNMP Management Frameworks”, Apr. 1999. [RFC99ter] RFC 2574, U. Blumenthal, B. Wijnen, “User-based Security Model (USM) for version 3 of the Simple Network Management Protocol (SNMPv3)”, Apr. 1999. [RFC99quater] RFC 2613, R. Waterman, B. Lahaye, D. Romascanu, S. Waldbusser, “Remote Network Monitoring MIB Extensions for Switched Networks Version 1.0”, Jun 1999. [SAY96] Saydam, T., Magedanz, T., “From Networks and Network Management into Service and Service Management”, Journal of Networks and System Management Vol. 4, No. 4, pp. 345-348, 1996
[TAN96] Tanenbaum, A., Computer Networks, Prentice Hall, 1996.
[TRU01] Trulove, J., LAN Wiring, McGraw-Hill, New York 2nd ed., 2001.
8. Test questions Question 1: Which of the following are used in order to evaluate a change impact? A) The effect of not implementing the Change B) The impact on non-IT infrastructures within the organisation C) The IT, business and other resources required to implement the Change D) Additional ongoing resources required if the Change is implemented
Answer A) a, b, c, d Right
B) a, b No, correct answer is: a, b, c, d
C) a, c No, correct answer is: a, b, c, d
D) b, c, d No, correct answer is: a, b, c, d
Question 2 Which are main benefits of Change Management process? A) improved risk assessment activities B) a reduced adverse impact of Changes on the quality of services and on SLAs C) better assessment of the cost of proposed Changes before they are incurred D) a reduced number of CIs involved in RFCs
Answer A) a,b,d No, correct answer is: a, b, c
B) a,c No, correct answer is: a, b, c
C) b No, correct answer is: a, b, c
D) a,b,c Right
Question 3 What does MIB deal with? A) MIB stores inside itself the solutions employed for fixing the failures occurred within the network. NO, MIB is a database in which merge the information gathered by the agents monitoring the components
B) MIB holds the application logic for fixing the failures within the network NO, MIB is a database in which merge the information gathered by the agents monitoring the components
C) MIB is a database in which merge the information gathered by the agents monitoring the components Right
D) MIB holds the information sent by NMS NO, MIB is a database in which merge the information gathered by the agents monitoring the components
Question 4 Network administrator can monitor users and devices access to network resources by: A) Performance Managing NO, Performance Managing aims at quantifying, measuring, report making, analyzing, and controlling different network devices performances. Correct answer is Account Managing.
B) Security Managing NO, Security Managing aims at establishing and respecting control indication for network resources access. Correct answer is Account Managing.
C) Configuration Managing NO, Configuration Managing enables network configuration data and devices monitoring, as to control and manage malfunctioning effects on hardware and software element composing the network. Correct answer is Account Managing.
D) Account Managing Correct
Question 5 Network managing protocol Notification functionality enables: A) the agent to communicate to central managing station the occurrence of important events concerning the monitored device. Correct
B) the managing station to recover the received data value housed within MIB NO, it enables the agent to communicate to central managing station the occurrence of important events concerning the monitored device.
C) the managing station to set values within MIB NO, it enables the agent to communicate to central managing station the occurrence of important events concerning the monitored device.
D) the managing station to recover and set data values housed within MIB NO, it enables the agent to communicate to central managing station the occurrence of important events concerning the monitored device
Question 6 Which message type in SNMPv2 does not need a response by the receiving unit? A) GetRequest NO, correct answer is TRAP
B) InformRequest NO, correct answer is TRAP
C) GetBulkRequest NO, correct answer is TRAP
D) TRAP Correct
Question 7 Which is the main difference between protocols SNMP and SNMPv2? A) SNMPv2 is based on TCP transport level protocol while SNMP is based on UDP protocol NO, both of them are based on UDP transport level protocol. Correct answer is: SNMPv2 protocol supports a centralized network managing as well as a distributed network managing while SNMP supports only the former.
B) SNMPv2 is based on UDP transport level while SNMP is based on TCP protocol NO, both of them are based on UDP transport level protocol. Correct answer is: SNMPv2 protocol supports a centralized network managing as well as a distributed network managing while SNMP supports only the former.
C) There is no difference between the two protocols NO, SNMPv2 protocol supports a centralized network managing as well as a distributed network managing while SNMP supports only the former.
D) SNMPv2 protocol supports a centralized network managing as well as a distributed network managing while SNMP supports only the former. Correct
Question 8: Protocol SNMP version 3 was introduced in order to outperform the limits in previous versions. In particular, the more significant and innovative contributes are in function:
A) Managing and Security Correct
B) Managing NO, protocol SNMPv3 also introduces functionalities regarding network security.
C) Security NO, protocol SNMPv3 also introduces functionalities regarding network managing
D) Managing and control NO, protocol SNMPv3 introduces functionalities regarding network managing and security
Question 9: Malfunctioning management is a process involving the following fundamental steps:
A) Components checking and identification, and checking agents conduct definition. NO, the right steps are: malfunctioning identification and classification – isolation of its cause – fault correction strategy selection and accomplishment
B) Fault causing component identification and classification NO, the right steps are: malfunctioning identification and classification – isolation of its cause – fault correction strategy selection and accomplishment
C) Spreading time control and racket transmission NO, the right steps are: malfunctioning identification and classification – isolation of its cause – fault correction strategy selection and accomplishment
D) The right steps are: malfunctioning identification and classification – isolation of its cause – fault correction strategy selection and accomplishment Right