About the IMMS Framework

 What is Maintenance Management?

Maintenance Management is a frame work type service employed by many industries today such as the automotive, oil refineries, and factories. It is hence interpreted differently by different industries and according to the particular industry’s individual requirements.


The Maintenance Management System delivers a service where its fundamental model is the primary means for understanding the major functions of network Maintenance Management Systems. This model consists of two major schools and four conceptual areas: Preventive Maintenance and Corrective Maintenance schools.

The Models then consists of the following five conceptual areas of Information Management:

Traffic-Flow Maintenance Management

The goal of Traffic-Flow maintenance management is focused on maintaining the Flow of management information traffic on the Network; with an emphasis on avoiding Management information traffic storms across LAN and WAN links between NMS requests and the Network Element’s responses to those requests. 

Flow Maintenance involves three main steps:

First, understanding and specifying the information to be collected from which of the managed network elements.

Second, bundling of the requests to be sent from the NMS and determining by which of the managed Network Elements (or groups thereof) first the bundled requests to be received.
Third, time lining the bundled requests so as not to congest the Network links giving the devices the time to respond to requests in a timely manner. 
Management Requests should be mainly an SQL SELECT type of operation; where the targeted element’s relevant information is targeted without having to continuously use GetNext or GetBulk operations which require a dedicated and high cost search criteria, one that is present in INDEX components.

Historically; NMS generated management queries have been launched uncontrollably (using BULK operations for example) to all elements detected in the network, prompting each and every element to reply back to every request. This has been known to overwhelm NMS Servers as well as managed network element’s resources significantly enough to cause link congestions. This indeed has been known to cause the loss of the very management information itself that the NMS is originally trying to collect.

Allowing Network Administrators to logically Bundle sets of large and extensive network management data requests into a number of autonomously controlled management processes, may it targets information directly related to the physical elements themselves, or a certain service implemented on those elements, or a subset of elements, or indeed a subset of one or more of the services running on those elements; is essential to allow for a full scalability of management traffic-flow.

Authority Maintenance Management

The goal of the Authority maintenance management is the maintenance of authoritative rights and ownerships over managed information. 
Authority over managed information is exercised by means of an application User(s) with the appropriate rights to initiate pre-determined, planned, and structured set of management information requests to be sent to a pre-determined, planned, and structured set of the network element(s) to be managed.

Another implicit and equally important goal for exercising this Authority over the managed information is the grouping of managed network elements into logical entities to allow the NMS to forward its intended data request sets to each managed group independently from other groups, where the particular set of requests could be considered unnecessary overhead. 

Allowing network administrators to logically break large networks into a number of autonomously managed bodies, may it represents the physical elements themselves, or a certain service implemented on those elements, or even a subset of the total elements, or indeed a subset of one or more of the services running on those elements; is essential to allow for a full scalability of management solutions implementations.

Quality Maintenance Management

The goal of quality maintenance management is to maintain the relevant information that is considered of value to an organization. This should be evaluated according to the particular business objectives of the concerned organization combined with the management policies and technical requirements of that organization.

The key to advancing this quality of the collected and acquired information is by: Building a layered structure of management information to be derived from the understanding of existing network information resources. Knowing what is valuable according to a pre-planned structure of network information resources should lead to the exclusion of maintaining less valuable information. Un-managing less valuable information will have the added value of freeing more resources towards maintaining valuable information while alleviating the shortcomings caused by spending valuable resources on managing less interesting and less valuable information.

It is essential that the decision of what information to manage be totally up to the concerned organization that owns the information resource pool. This is a significant shift from the traditional vendor driven management solutions where organizations would have to adhere to the vendor’s solutions and only decides on the outcome and whether the results are useful or not. By this time of course management resources have already been expended.

Two motivations should drive the Quality of managed information to be collected by a particular organization. One is by solving a business problem or increase efficiency of the current methods of accomplishing work while reducing overall costs. A solution that doesn`t save money while providing better services; isn`t probably worth accomplishing. The other is for the Quality of managed information to be structured around the real work processes within the organization. Hence it should focus the tools toward those staff members supporting the managed area in a manner which makes their job easier and faster.

Data Integrity Maintenance Management

The goal of the Integrity maintenance management is to look for and to implement ways to preserve the collected data for as long as it is necessary and indeed useful as per the organization’s requirements determined by the network administrators.

NMS systems databases typically and rapidly grow very consistently. The rapid growth of database volumes could deepen the gap between data generation and data understanding. Hence Summarization is necessary as it reduces a large number of actual database tuples into a relatively small number of generalized descriptions.

Summary data is manufactured by analyzing and manipulating detailed data to derive new technically refined (more meaningful) or business data. The data we capture about devices, such as their names, IP addresses, images running and traffic usage, is detailed data, as is the data required to record faults and maintain inventory, and respond to events. The data used to assess how well the device is performing, such as link utilizations, and resource usage per running services is in nature summary data.

Summary data often represents the aggregations or counts of detailed data across important technology or business hierarchies. Historical data represents the state of the network at specific points in time. Consequently, historical data is non-volatile. Once the snapshot has been taken, the values are never changed, so that the image of the network reflected in the snapshot is never lost.

Snapshots are taken regularly, such as on the last business day of the month, which allows the network administrators to compare the values in different snapshots. Some of the snapshots could be designated as baselines and stored as is without any additional summarization to enhance the value comparisons. Future data is a projection of how the network will look at a future point in time.

A network may have a number of different projections for the same future time frame to represent various scenarios or topologies maybe. Each projection is based on a set of assumptions about how the network environment will evolve. For example, you may develop alternate projections to study the impact of different link utilization growth rates on the VLAN network design topology. This future data is used as a metric against which historical data is analyzed.

Data sharing could be the direct result of distribution of data management functions across the network. Data sharing has always been a major concern for data management professionals battling the disparate data issues associated with locally developed applications. The data requirements framework of IMMS systems helps to highlight the need to implement data management practices that satisfy the data sharing needs of each network segment or service.

Network topology is a major determinant of how data could be distributed across network environment. A multinational organization may strongly believe that each business segment must be able to operate as an autonomous entity and may adopt a technology architecture based on data held in local databases. The data management issues involved in implementing a data sharing strategy for this distributed data environment are more challenging than for an organization that decides to implement a centralized data management strategy for current and historical detailed data, while allowing local management for the corresponding summary data.

Collected data from the managed network elements are either reported directly in real time. Or it could be stored into a repository to be viewed at specific times (such as data intended for billing or accounting). Real time data could also be viewed and discarded, or viewed and kept for historical reporting. Depending on the managed information and its value to the organization; real time data should be kept for some reasonable period of time to allow for an accurate view of the data and then summarized gradually over controlled periods of time to keep a reasonable view of the data and a meaningful history of the network conditions.

The overall data whether the real time or the summarized part should be backed up by the NMS system to be retrieved whenever –or if- a catastrophic event shall compromise the integrity of the data. Two types of backups should always be considered by the management system; one that stores a recent copy of the data replacing the old data.

Another that stores a copy of recent data incrementally by keeping the older backup copies intact. Rapidly increasing amounts of data motivate the need for compressing data stored in databases. This might be a necessary step if growth levels forecasts compromise performance of host servers.

Presentation Maintenance Management

The goal of presentation maintenance management is to focus on the reporting of the collected information from the management resources pool of a particular Network.

Viewable reports have to adhere to acceptable set of norms to reflect the validity, the integrity, and readability of the relevant information representing the business objectives, and management policies of a particular organization.

Reporting type and style should be dependent on the syntax data type and syntax structure of each of the managed pieces of information to be reported. Management information should then be classified according to the Data types of the objects to be reported and then according to the ‘true meaning’ of the value of the object in the context of the network resource being managed.

Reported data information (or managed objects) could be any one of the following: A directly acquired object from the Managed Element, or an indirectly acquired object from learning about an Managed Element (Learned object), or in fact calculated by combining two or more of those Elements (Calculated object). Furthermore, Reported objects could also be directly solicited by a Management Process or conditionally solicited; the later is achieved by acquiring a managed object directly and testing its value first before deciding on obtaining other objects.

It is important to mention here that not all managed object (whether directly or indirectly acquired) have to be reported. Such are those objects that have to be acquired only for their values to be tested against a certain criteria, before acquiring other objects that in turn may or may not be reported according to the particular organization’s requirements. Information associated with a problem or symptom should mean something to the support personnel. If they see the problem at a glance, they should know to which specific area the problem belongs and what to do to get started in the trouble isolation process.

Reported Objects should not report object values that are considered erroneous. Individual Managed Objects output value limitations should be understood and certified by the NMS’s Management Process before being reported. This is necessary as historically a single wrong value could throw a graph off scale and deny the administrators the chance to view what could be hundreds maybe thousands of data points that are valid. 

Maintenance Management in a nutshell. The Integrated maintenance of the Flow of management information, the Authority over the information, the quality of the information, The Integrity of the information, and finally the Presentation of the information to varying degrees in any of all the above areas in either the Preventive or the Corrective sense, is really the focus of the next era/generation of Network Management. IMMS is the next layer of services to complement the NMS model.