/
Patchwork is a manufacturing and distribution company with headquarters in Cleveland, Ohio. Patchwork is currently evaluating desktop management suite products to manage their install base of over 10,500 users located mainly in their Cleveland and Detroit locations, with a smaller percentage of users spread across 40 national sales sites. Necessary features of a desktop management suite, as indicated by Patchwork, include hardware and software inventory, software distribution, remote control, and hardware configuration capabilities. The successful management product will be a tightly integrated solution that shares a common database. Support for a web-enabled interface is also necessary. The business requirements that Patchwork is looking to address by implementing a desktop management suite include an annual 15% reduction in the total cost of owning and maintaining their PC infrastructure. This is best effected by application standardization through software distribution, remote troubleshooting and repair through remote control, and accurate asset forecasting though reliable asset and inventory data. Future plans include the implementation of a directory service for centralized user administration and policy-based management. Overview Network management via SNMP has evolved greatly over the past decade to keep pace with the rising demands that organizations place on their IT departments. IT administrators have gone from wanting to know when the network is experiencing trouble to when the network will experience trouble, all due to the efficiency of organizations in mapping their business requirements to technology. This has put great pressure on network management products to evolve to meet the ongoing needs of the IT staff. Now network management is simply not enough. Over the past few years the industry has seen management products begin to reach past managing networking hardware and begin to touch the desktop PC. At first, it was acceptable to address these new challenges with disparate point solutions that didn't tie back into an overall management framework. But now with the fiercely competitive e-business marketplace, an organization needs to find an advantage anywhere possible. The efficient communication of critical information between IT and executives becomes dependent upon the efficient sharing of critical data between the tools that IT depends on. Aprisma Management Technologies, www.aprisma.com, has been developing SNMP-based management software since shipping its first version of SPECTRUM Enterprise Manager in 1991. Aprisma has had a front row seat during the evolution of network management during the 90's and has led much of the way. SPECTRUM introduced the use of an Object Oriented Database with version 1.0 in 1991 and with SPECTRUM's patented Inductive Modeling Technology, Aprisma truly has been an innovator in the field of SNMP-based management technology. Aprisma will team with Metrix s.a.r.l., the fastest growing system management vendor focused on desktops and servers worldwide, to offer a deeply integrated solution that will scale to meet the needs of an enterprise the size of Patchwork's. Metrix, www.metrixsystems.com, is an independent software development company based in the Telecom Valley in Sophia Antipolis, France, the biggest Research & Development Park in Europe. Metrix serves the market worldwide with the goal of creating best-of-breed desktop and systems management solutions for medium-to-large scale corporate networks. Metrix was created in 1994, with the mission of developing best-of-class management solutions addressing the problem of managing desktop and server systems in medium to large-scale enterprise networks. In 1995, Metrix commercialized its first version of WinWatch to the corporate market. Metrix was recently elected the Most Innovative Start-up of the Year by The Enterprise Forum. For more information, please visit www.metrixsystems.com/news /pressreleases/startup.htm. Proposed Solution The proposed solution uses SPECTRUM Enterprise Manager to take into account that management is more than just desktop management. While Patchwork is focused on Desktop management today, SPECTRUM will provide the framework from which to integrate other aspects of management tomorrow. From the core platform of SPECTRUM, Metrix WinWatch will utilize data gathered from intelligent agents installed on managed desktops to perform managed functions such as ESD, Remote Control, Asset Management, Change Management, and HW/SW inventory tracking (from a single WinWatch agent per PC). The solution leverages SPECTRUM's object oriented design with the WinWatch intelligent agents. The tight integration between the two products creates a solution that features industry leading fault isolation capabilities, self healing agents that negate the need for multiple agents on a single platform, and the capabilities to perform a robust array of management tasks. As an innovator in Web technology, Metrix has also developed a Web browser management console for SPECTRUM, enabling full-service, browser-based management. The Metrix Web Console for SPECTRUM includes innovations such as multiple language support, user preferences, and secure access. WBEM and CIM capabilities will be gained through the inclusion of the Metrix Web Console and the SPECTRUM Web Operator application, which has a certified Cisco Management Connection. Cisco Systems, Inc. is the world wide leader in networking for the Internet. For more information please visit www.aprisma.com/ournews/ 1999/dec/12-15a.html. No solution is complete with out the provision of support services and training. Aprisma Management Technologies has included 7X24 technical support and free product upgrades with the overall project costs. To fully realize the benefits of the power of SPECTRUM and WinWatch, Aprisma recommends training that can be held either on-site or through Aprisma's world Wide training Center located in Portsmouth NH.
Financial Summary
Business Process Management
Leading information businesses are seeking to integrate their supply chain and business processes both internally with their business units and employees and externally with their suppliers and partners through the use Enterprise Resource Planning tools like BAAN, Oracle, and SAP R/3. An important step in planning for the installation of these ERP systems is the re-definition of business processes within the organization. Here is where leading professional services consulting firms such as Andersen Consulting, Ernst & Young, KPMG and Price Waterhouse Coopers help the business to define new processes to improve efficiency while also enabling new access to services by both internal and external partners, customers and employees.
SPECTRUM and the CONTINUITY advanced application enable your company to take the business process flow charts developed in conjunction with your consultants, and directly map these processes with associated icons directly into SPECTRUM's KnowledgeBase. Leveraging Continutiy, SPECTRUM understands how the people, departments, applications, computers, and networking equipment impact and are impacted by the various business processes. CONTINUITY is an IT service management solution for the global and virtual enterprise, designed to help IT organizations manage service requirements in complex distributed computing environments. CONTINUITY allows organizations to manage, control and monitor availability, service levels and operational risks for all IT resources-networks, systems, databases and applications-in terms that both IT operations managers and business managers can understand. To maximize availability and minimize downtime, CONTINUITY helps managers to correct and prevent service disruption by identifying the root cause and impact of a service failure. Using a self-describing model of the enterprise, CONTINUITY offers a highly scalable solution designed to meet the needs of a Fortune Global 1000 organization. For more information on SPECTRUM and CONTINUITY, please visit www.aprisma.com/partners/featured/ics.html.
In December 1997, Data Communications Magazine reported that 67% of the Information Technology managers polled would be implementing Service Level Agreements by the end of 1999. Today, application service providers (ASPs), Internet service providers (ISPs), and internal IT organizations are all being challenged to deliver some guarantees around the services that they offer. Commerce is increasingly reliant on services that are delivered by information technology. Service owners responsible for the management of information technology walk a tightrope, balancing demands from service consumers and the costs of service provision while ensuring that the enterprises' information technology service expectations are met.
Service Level Agreements are contracts between service consumers and service owners or providers that set the parameters of the service. The complexity of information technology, the rapid pace of change, and increasing demands from, and sophistication of, service consumers make it difficult to manage service levels effectively. Without good tools, Service Level Agreements become yet another piece of paper. A good Service Level Agreement management application provides the means to manage your business more effectively. Specifically, a Service Level Management solution should provide multi-vendor capabilities, which scale to meet the global requirements of your company by measuring:
www.aprisma.com/white-papers/spectrum/slm.pdf Another key function of Business Process Management is the ability to provide accurate and timely reports to business managers about the IT resources they rely upon to compete. SPECTRUM's advanced application Executive Information System (EIS) is tightly integrated with the SPECTRUM KnowledgeBase to provide the consolidated, executive-level analysis and reports you need to make the right decisions for your business. When used with SPECTRUM, EIS provides web-based reports in a concise format allowing executives to easily understand how the network infrastructure is supporting their company's information technology needs. EIS automates the reporting process, eliminating the need to manually run reports or manage archived data. It's like having your own administrator whose sole function is to organize network data and provide you with the information needed to meet your management objectives. If you are responsible for the cost-effective performance of the network and technical infrastructure on which your enterprise and your customers depend, the last thing you need is more "raw" information. What you need is an easier, faster way to consolidate, analyze and interpret the vast amount of data that Spectrum has already collected. EIS presents a modular architecture which allows for the development and deployment of field-installable modules that can be tailored to meet various reporting requirements including:
Artificial Intelligence
The foundation of SPECTRUM is Inductive Modeling Technology (IMT), influenced by concepts and techniques developed in various fields of artificial intelligence, such as linguistics, knowledge representation, and problem-solving. Although it can be categorized as either a semantic network or a model-based system, IMT integrates the concepts and technologies of both.
The core of IMT is a knowledge-base which represents everything modeled by SPECTRUM. The knowledge-base consists of all defined concepts stored in the database and contained within the procedural knowledge of inference handlers in SpectroSERVER. SpectroSERVER provides the interfaces to access this information.
The knowledge-base is populated with concepts, relationships, and specific instances of those concepts and relationships - including instances representing managed nodes, as well as protocols for managing the nodes. SPECTRUM can manage any nodes that use Simple Network Management Protocol (SNMP) and Internet Control Management Protocol (ICMP Ping). If a managed node does not use these protocols, a protocol converter can be developed using the SPECTRUM External Protocol API.
Model types, relations, and rules form the basic building blocks of the SPECTRUM knowledge-base. Model types and relations are abstractions of real-world objects, and rules are logical restrictions on the way these abstractions can be linked to each other. The concepts and relationships in the SPECTRUM knowledge-base are visually presented from different perspectives called views. For example, some views depict logical connections between managed nodes, while others display physical locations. Some of the views are hierarchical, enabling users to drill down to detailed sub-views. Other views are generic, containing gauges, graphs, and charts that monitor the activity of a managed node. The instances of the concepts in the SPECTRUM knowledge-base are shown as icons in various views. Each icon is a "live" picture of a model, some containing status fields that change color or update in real time as the data in the knowledge-base changes.
SpectroSERVER Database
SpectroSERVER contains an object-oriented database. The database provides storage for specific device configurations, statistics and events, and a modeling catalog of model types and relations.
The Virtual Network Machine (VNM) contains procedural knowledge, or SPECTRUM intelligence, in the form of inference handlers that describe how a model type reacts to changes in its environment. Several inference handlers may be required to define how a model type should react. This group of inference handlers is referred to as an intelligence circuit, one or more of which define the behavior of the model type.
The VNM receives requests from the SpectroSERVER API and calls on the appropriate component to service the request. The Device Communication Manager (DCM) is the interface between the VNM and the managed nodes. The DCM includes various clients that communicate with managed nodes via their protocols. There is one client for each protocol supported.
The SPECTRUM model type hierarchy is a collection of model types that exist in the SPECTRUM knowledge base. The model types are built in a hierarchical fashion, with more specific model types being derived from more general ones. When a model type is derived from a general one, it inherits both the attributes and the intelligence circuits from the base model type. The derived model type also participates in all the relations in which the base model type participates.
Seven types of entities are recognized in SPECTRUM:
1. Device is any electronic device that can respond to commands from SPECTRUM, either directly via a resident management agent or a proxy agent, or via any number of other devices. It forms the base of devices such as routers, bridges, hubs, and workstations that are managed directly by SPECTRUM, as well as devices that can only be managed indirectly, such as multi-port transceivers.
2. Organization represents organizational entities, such as Enterprises, Departments, and Divisions. This model type is used to show ownership of devices.
3. Location is where a managed node is placed at a particular time, such as Country, Building, Room, or Rack.
4. Component represent hardware components (Boards or Ports) and software components (logical groupings, such as a group of ports on a hub).
5. Topology is the logical layout of a collection of managed nodes.
6. SoftwareApp is any software application managed by SPECTRUM.
7. Some model types are combinations of some of these entities. Links or segments of cables, known as Fanouts, are physical representations of network topologies. As such, they can be classified as Devices, as well as Topologies and Components. These model types are grouped under Link.
It is through this modeling intelligence that IMT is capable of inferring the condition of every device on the network, enabling the VNM to isolate faults to the device level by understanding the interaction of network components. When a fault is detected by the VNM, the Artificial Intelligence engine is able to infer that the downstream devices cannot be polled due to a failure of the upstream device. SPECTRUM intelligently filters and suppresses alarms from devices downstream of the fault, only passing through an alarm for the faulty device.
Distributed Management
Distributed SpectroSERVER (DSS) is a modeling feature that uses the concept of landscapes. A landscape is composed of the models, associations, attribute values, alarms, events, and statistics specific to any one SpectroSERVER. Using DSS, networks can be logically partitioned into subnets or multiple landscapes, each with its own local SpectroSERVER. Landscape icons represent distributed server databases, and provide a rollup of alarm information for the devices modeled in those remote databases. Each remote landscape is identified by a unique landscape handle. DSS can improve SPECTRUM performance when managing a large network by distributing the load introduced by management traffic and delegating network management functions to remote workstations.
When a network is modeled with multiple landscapes, it is possible for a single SpectroGRAPH (or other client application) to access information from more than one SpectroSERVER. SPECTRUM's Distributed Data Manager provides access to data from all the SpectroSERVER databases in the managed environment. In addition, applications are distributed. Rather than requiring an application on each server, a single instance of an application can be used across multiple domains, enabling consolidation of alarm monitoring, reporting, inventory, and other key management tasks across the entire enterprise from a single console.
Conversely, the use of multiple SpectroGRAPHs enables division of the IT environment into management domains based on geography, job function, or device type. It also enables an IT organization to distribute administrative responsibility or centralize it in a "control center."
The diagram [click here to see diagram] illustrates SPECTRUM's client/server architecture, as well as device and systems management components of an enterprise solution.
Fault Tolerance (Fail-over Capability)
SPECTRUM's solution for redundancy is based on primary and secondary SpectroSERVERs, also called Fault Tolerance. Having more than one SpectroSERVER managing a landscape provides Fault Tolerance. At any time, only one copy of a landscape is active. The predominantly active landscape is designated the primary SpectroSERVER. The inactive copies are secondary or tertiary SpectroSERVERs. If a secondary SpectroSERVER fails, the primary can be set up to register an alarm that notifies the operators that redundancy has been lost. A secondary SpectroSERVER must be dedicated one-to-one with the primary and mirror its device management modules. When a primary SpectroSERVER fails, a secondary SpectroSERVER becomes active and starts managing the network. Any client applications connected to the primary SpectroSERVER are automatically switched to the secondary SpectroSERVER. When the primary comes back, the applications are automatically switched back to the primary, and the secondary becomes inactive.
Fault tolerance features can be divided into three areas:
Frame Relay is rapidly growing in popularity as an economical and efficient method of connecting offices, especially linking remote offices to corporate headquarters. The benefits of the Frame Relay protocol include lower cost than dedicated T1 lines and multiple connections on a single interface, which allows companies to save a considerable amount of money on routing hardware. As the Frame Relay market continues to grow, so does the need to have real-time information and reports available for each Permanent Virtual Circuit (PVC) in a network. Cabletron offers a scalable Frame Relay management application, fully-integrated with SPECTRUM, that is designed for the Internet Service Providers (ISPs) and corporate user markets. SPECTRUM Frame Relay Manager
SPECTRUM Frame Relay Manager allows Frame Relay carriers, their customers and corporate users to enjoy a greater level of information than has been available from the industry providers in the past. It gives users precise monitoring of Frame Relay circuit bandwidth usage and costs with the use of its detailed reports on bandwidth utilization, line congestion and performance monitoring. Users can easily customize reports to show real-time information at specified time intervals. Features
SPECTRUM Frame Relay Manager can monitor any Frame Relay Access Device (FRAD) that supports the standard IETF Frame Relay MIB (RFC 1315) for SNMP data collection. Frame Relay Manager automatically creates an application model representing management support of Frame Relay firmware. Management support of Frame Relay firmware is defined by the RFC 1315 MIB. This MIB contains three groups: one defines the Data Link Connection Management Interface (DLCMI), one defines the circuits, and one describes errors. In addition, one trap is defined which indicates changes to a virtual circuit. The DLCMI table is used to specify various monitoring parameters, such as polling interval and error threshold, and to set the maximum number of virtual circuits. The Circuit table provides information about each virtual circuit for statistics reporting and performance monitoring. The Error table records error information for each virtual interface. The three tables of this MIB can be displayed from this application model. AutoDiscovery
SPECTRUM Frame Relay Manager provides enhanced AutoDiscovery. This maps the Frame Relay network and resolves connections on the DLCMI-to-DLCMI level. A new model type - FrameRelayDTE - visible in the topology view, shows connections of active Frame Relay interfaces, and contains a set of PVCPort models to represent each of the PVC connections. A PVC model is used to indicate that a pipe represents a Frame Relay connection. Alarms
SPECTRUM alarms are generated and users are notified immediately when load and congestion thresholds are exceeded, or in the case of circuit failure. SPECTRUM monitors each PVC for congestion and allows users to define thresholds for throughput. SPECTRUM's fault isolation will identify a down PVC even if the remote router is not modeled. Alarms are generated in response to the MIB RFC 1315 trap that indicates changes in a DLCMI and its current state. Users are notified of status changes before the next poll. Reporting
SPECTRUM Frame Relay Manager provides statistics per PVC at user-defined intervals that are presented in customizable reports, eliminating dependency on reports provided by carriers. These reports can be used to determine a need for greater bandwidth, or to identify circuits with CIR greater than current need.
Aprisma Management Technologies
ClickNet Software
Intel
Microsoft
Vector Networks
Desktop management suite RFP - Aprisma Management Technologies
EXECUTIVE SUMMARY
Current EnvironmentPatchwork is a manufacturing and distribution company with headquarters in Cleveland, Ohio. Patchwork is currently evaluating desktop management suite products to manage their install base of over 10,500 users located mainly in their Cleveland and Detroit locations, with a smaller percentage of users spread across 40 national sales sites. Necessary features of a desktop management suite, as indicated by Patchwork, include hardware and software inventory, software distribution, remote control, and hardware configuration capabilities. The successful management product will be a tightly integrated solution that shares a common database. Support for a web-enabled interface is also necessary. The business requirements that Patchwork is looking to address by implementing a desktop management suite include an annual 15% reduction in the total cost of owning and maintaining their PC infrastructure. This is best effected by application standardization through software distribution, remote troubleshooting and repair through remote control, and accurate asset forecasting though reliable asset and inventory data. Future plans include the implementation of a directory service for centralized user administration and policy-based management. Overview Network management via SNMP has evolved greatly over the past decade to keep pace with the rising demands that organizations place on their IT departments. IT administrators have gone from wanting to know when the network is experiencing trouble to when the network will experience trouble, all due to the efficiency of organizations in mapping their business requirements to technology. This has put great pressure on network management products to evolve to meet the ongoing needs of the IT staff. Now network management is simply not enough. Over the past few years the industry has seen management products begin to reach past managing networking hardware and begin to touch the desktop PC. At first, it was acceptable to address these new challenges with disparate point solutions that didn't tie back into an overall management framework. But now with the fiercely competitive e-business marketplace, an organization needs to find an advantage anywhere possible. The efficient communication of critical information between IT and executives becomes dependent upon the efficient sharing of critical data between the tools that IT depends on. Aprisma Management Technologies, www.aprisma.com, has been developing SNMP-based management software since shipping its first version of SPECTRUM Enterprise Manager in 1991. Aprisma has had a front row seat during the evolution of network management during the 90's and has led much of the way. SPECTRUM introduced the use of an Object Oriented Database with version 1.0 in 1991 and with SPECTRUM's patented Inductive Modeling Technology, Aprisma truly has been an innovator in the field of SNMP-based management technology. Aprisma will team with Metrix s.a.r.l., the fastest growing system management vendor focused on desktops and servers worldwide, to offer a deeply integrated solution that will scale to meet the needs of an enterprise the size of Patchwork's. Metrix, www.metrixsystems.com, is an independent software development company based in the Telecom Valley in Sophia Antipolis, France, the biggest Research & Development Park in Europe. Metrix serves the market worldwide with the goal of creating best-of-breed desktop and systems management solutions for medium-to-large scale corporate networks. Metrix was created in 1994, with the mission of developing best-of-class management solutions addressing the problem of managing desktop and server systems in medium to large-scale enterprise networks. In 1995, Metrix commercialized its first version of WinWatch to the corporate market. Metrix was recently elected the Most Innovative Start-up of the Year by The Enterprise Forum. For more information, please visit www.metrixsystems.com/news /pressreleases/startup.htm. Proposed Solution The proposed solution uses SPECTRUM Enterprise Manager to take into account that management is more than just desktop management. While Patchwork is focused on Desktop management today, SPECTRUM will provide the framework from which to integrate other aspects of management tomorrow. From the core platform of SPECTRUM, Metrix WinWatch will utilize data gathered from intelligent agents installed on managed desktops to perform managed functions such as ESD, Remote Control, Asset Management, Change Management, and HW/SW inventory tracking (from a single WinWatch agent per PC). The solution leverages SPECTRUM's object oriented design with the WinWatch intelligent agents. The tight integration between the two products creates a solution that features industry leading fault isolation capabilities, self healing agents that negate the need for multiple agents on a single platform, and the capabilities to perform a robust array of management tasks. As an innovator in Web technology, Metrix has also developed a Web browser management console for SPECTRUM, enabling full-service, browser-based management. The Metrix Web Console for SPECTRUM includes innovations such as multiple language support, user preferences, and secure access. WBEM and CIM capabilities will be gained through the inclusion of the Metrix Web Console and the SPECTRUM Web Operator application, which has a certified Cisco Management Connection. Cisco Systems, Inc. is the world wide leader in networking for the Internet. For more information please visit www.aprisma.com/ournews/ 1999/dec/12-15a.html. No solution is complete with out the provision of support services and training. Aprisma Management Technologies has included 7X24 technical support and free product upgrades with the overall project costs. To fully realize the benefits of the power of SPECTRUM and WinWatch, Aprisma recommends training that can be held either on-site or through Aprisma's world Wide training Center located in Portsmouth NH.
Proposed Solution Feature Matrix | Feature |
Yes |
No |
Future |
| Electronic Software Distribusion | X | ||
| Remote Control | X | ||
| Server Management | X | ||
| Asset Management | X | ||
| Security Management | X | ||
| Enterprise Management | X | ||
| Directory Enable Services | X | ||
| Software Cost | |
| $50,000.00 | |
| $15,000.00* | |
| $588,000.00 | |
| $29,400.00 | |
| $667,400.00 | |
| Maintenance Cost | |
| $120,132.00 | |
| Total Solution Cost | $787,532.00 |
SPECTRUM ENTERPRISE MANAGER
Introduction The Internet is changing how we work, live and play...just about everything. The new world economy is being driven by the explosive growth of the Internet and the new forms of business-to-business electronic commerce it enables. In many ways, the next industrial revolution is at hand. Business is changing to keep pace. In fact, to survive, businesses of all sizes are rapidly becoming information businesses. Information Business can make small companies look like big companies as it extends the reach of the organization globally to deliver customer service anywhere, anytime, anyhow. Information Business also enables individualized, interactive, and immediate communications with your customers as you seek to deliver millions of customized solutions targeted at one customer rather than one solution for all. Mass Customization via e-mail, rather than direct mail, can provide focused, value-oriented experiences that boost customer loyalty as E-commerce is replacing traditional brand management. Your customer is just a click away from doing business elsewhere so it is imperative that you ensure an optimal encounter by each and every customer. Information...get IT...move IT...use IT. Information Technology (IT) has taken center stage as the critical asset to ensure future competitive advantage. Linking IT with an organization's business strategy has become a fundamental requirement. Traditionally viewed as a cost center, IT has evolved to become a profit enabler. Strategic business processes that define near-term competitiveness include Electronic Commerce, Electronic Resource Planning, and Knowledge Management. IT is all about getting the right information to the right people at the right time. However, managing information business change cannot be the sole responsibility of IT technologists. IT should be viewed as a Service, measurable in terms of its reliability and its ability to grow with the business. There is a tremendous need to understand how IT impacts, and is impacted by the Business and people it is designed to support. The Business of IT must offer benchmarks, best practices, and tools to gauge the effectiveness of investments through all phases of planning, implementation and measurement while providing increased security, reliability and asset utilization for business-critical applications. The challenge of IT complexity quickly arises, however. Reliability has still not delivered "Five Nines" availability. "Five Nines" means that downtime can only be measured in seconds, not minutes per month. Complete failures are also no longer the main issue as IT service slow-downs known as "brown outs" are occurring twice as frequently as "black outs". A 1999 study from Infonetics Research conservatively estimated that network downtime and service degradations cost the average enterprise $3.9 million per year in lost revenues and productivity. Operational costs are escalating and technical complexity is growing administrative burdens. In fact, a recent study found reactive tasks, essential to keep the network up and running, but guaranteed to at best only maintain cost and service quality status quo in the face of increasing complexity and demand, account for 62% of total support staff time. How will you overcome the technical complexities to transform your business into an information-based E-Business? Keep your fingers crossed? Hire an army of technologists? Use technology to solve the technology problem? Both Deloitte Touche and Forrester Research estimate that there will be trillions of dollars of business to business electronic commerce conducted over the Internet within the next five years. What should be gleaned from that information is that no e-commerce can happen without reliable application, computing and networking IT infrastructures. Fundamentally, there is no e-commerce unless the reliability and operations of the networking infrastructure can be assured. Businesses are demanding more from IT departments today than they can deliver by simply installing a loose collection of software point products to separately manage the application, computing and networking foundations. Business managers want guarantees for their critical IT services. SPECTRUM...Managing Information Business. Aprisma Management Technologies has over a decade of experience managing some of the largest and most complex multi-vendor IT infrastructures in the world. While our company name, Aprisma, can be closely associated with our primary product offering SPECTRUM, there is more to the story. Our focus centers around the needs of our customers as they seek to intelligently deliver accelerated, proactive responses to IT service degradations or outages based on the service management architecture of SPECTRUM.a Accelerated p Proactive r Responsive i Intelligent s Service m Management a ArchitectureSPECTRUM's capabilities are both broad and deep including:
- Business Process Management
- Service Level Management
- Executive Reporting and Measurement
- Application Management
- Response Time and Availability
- Systems Management
- Servers
- Desktops
- Network Management
- LAN
- WAN
- Cable Broadband
- Telephony
- Service Level Management
| DataCommunications Magazine | "Hot Products" | 1999, 1998 |
| InternetWeek | "Best of Breed" | 1999, 1998, 1997 |
| LAN Magazine | "Product of the Year" | 1999, 1998 |
| Network Computing Magazine | "Editor's Choice" | 1999, 1998 |
| "Well Connected Award" | 1997 | |
| Network World | "#1 Customer Satisfaction" | 1999 |
| Networld + InterOp | "Best of Show Finalist" | 1999 |
- Application Response Time and Availability
- Computing Systems Resource Utilization
- Network Throughput, Bandwidth and Latency
www.aprisma.com/white-papers/spectrum/slm.pdf Another key function of Business Process Management is the ability to provide accurate and timely reports to business managers about the IT resources they rely upon to compete. SPECTRUM's advanced application Executive Information System (EIS) is tightly integrated with the SPECTRUM KnowledgeBase to provide the consolidated, executive-level analysis and reports you need to make the right decisions for your business. When used with SPECTRUM, EIS provides web-based reports in a concise format allowing executives to easily understand how the network infrastructure is supporting their company's information technology needs. EIS automates the reporting process, eliminating the need to manually run reports or manage archived data. It's like having your own administrator whose sole function is to organize network data and provide you with the information needed to meet your management objectives. If you are responsible for the cost-effective performance of the network and technical infrastructure on which your enterprise and your customers depend, the last thing you need is more "raw" information. What you need is an easier, faster way to consolidate, analyze and interpret the vast amount of data that Spectrum has already collected. EIS presents a modular architecture which allows for the development and deployment of field-installable modules that can be tailored to meet various reporting requirements including:
- ExecutiveView
- AvailabilityView
- ServiceView
- AssetView
- CapacityView
- SystemsView
- they'll wait an average of just 20 seconds to see something on the page
- you have a maximum of 40 seconds to complete the home page display
- within 6 seconds the customer has decided to stay or leave the site
- the customer will use a maximum of 6 mouse clicks to find the desired information
- the maximum amount of time spent on a site is, on average, 12 minutes
- Software Distribution, Inventory, Auditing and Licensing
- Hardware Inventory
- Configuration Assurance and Change Audits
- Remote Control and Troubleshooting
- Policy-based user, file and printer management
- Virtual Support Agents which take proactive measures in response to exceeded thresholds
- Cable modems from Motorola and Nortel
- Set-top-boxes from Scientific Atlanta and General Instruments
- Broadband router/cable modem termination systems (CMTS) for Cabletron SmartSwitch Routers and Cisco's Universal Broadband Routers
- Optical and ATM backbones from market leaders including Cabletron, Cisco, Nortel, Marconi and Lucent
- Service level management with web-based customer views
- Application and server management
- Asset and inventory reporting
- Configuration management
- Accounting for capacity planning
Artificial Intelligence
The foundation of SPECTRUM is Inductive Modeling Technology (IMT), influenced by concepts and techniques developed in various fields of artificial intelligence, such as linguistics, knowledge representation, and problem-solving. Although it can be categorized as either a semantic network or a model-based system, IMT integrates the concepts and technologies of both.
The core of IMT is a knowledge-base which represents everything modeled by SPECTRUM. The knowledge-base consists of all defined concepts stored in the database and contained within the procedural knowledge of inference handlers in SpectroSERVER. SpectroSERVER provides the interfaces to access this information.
The knowledge-base is populated with concepts, relationships, and specific instances of those concepts and relationships - including instances representing managed nodes, as well as protocols for managing the nodes. SPECTRUM can manage any nodes that use Simple Network Management Protocol (SNMP) and Internet Control Management Protocol (ICMP Ping). If a managed node does not use these protocols, a protocol converter can be developed using the SPECTRUM External Protocol API.
Model types, relations, and rules form the basic building blocks of the SPECTRUM knowledge-base. Model types and relations are abstractions of real-world objects, and rules are logical restrictions on the way these abstractions can be linked to each other. The concepts and relationships in the SPECTRUM knowledge-base are visually presented from different perspectives called views. For example, some views depict logical connections between managed nodes, while others display physical locations. Some of the views are hierarchical, enabling users to drill down to detailed sub-views. Other views are generic, containing gauges, graphs, and charts that monitor the activity of a managed node. The instances of the concepts in the SPECTRUM knowledge-base are shown as icons in various views. Each icon is a "live" picture of a model, some containing status fields that change color or update in real time as the data in the knowledge-base changes.
SpectroSERVER Database
SpectroSERVER contains an object-oriented database. The database provides storage for specific device configurations, statistics and events, and a modeling catalog of model types and relations.
The Virtual Network Machine (VNM) contains procedural knowledge, or SPECTRUM intelligence, in the form of inference handlers that describe how a model type reacts to changes in its environment. Several inference handlers may be required to define how a model type should react. This group of inference handlers is referred to as an intelligence circuit, one or more of which define the behavior of the model type.
The VNM receives requests from the SpectroSERVER API and calls on the appropriate component to service the request. The Device Communication Manager (DCM) is the interface between the VNM and the managed nodes. The DCM includes various clients that communicate with managed nodes via their protocols. There is one client for each protocol supported.
The SPECTRUM model type hierarchy is a collection of model types that exist in the SPECTRUM knowledge base. The model types are built in a hierarchical fashion, with more specific model types being derived from more general ones. When a model type is derived from a general one, it inherits both the attributes and the intelligence circuits from the base model type. The derived model type also participates in all the relations in which the base model type participates.
Seven types of entities are recognized in SPECTRUM:
1. Device is any electronic device that can respond to commands from SPECTRUM, either directly via a resident management agent or a proxy agent, or via any number of other devices. It forms the base of devices such as routers, bridges, hubs, and workstations that are managed directly by SPECTRUM, as well as devices that can only be managed indirectly, such as multi-port transceivers.
2. Organization represents organizational entities, such as Enterprises, Departments, and Divisions. This model type is used to show ownership of devices.
3. Location is where a managed node is placed at a particular time, such as Country, Building, Room, or Rack.
4. Component represent hardware components (Boards or Ports) and software components (logical groupings, such as a group of ports on a hub).
5. Topology is the logical layout of a collection of managed nodes.
6. SoftwareApp is any software application managed by SPECTRUM.
7. Some model types are combinations of some of these entities. Links or segments of cables, known as Fanouts, are physical representations of network topologies. As such, they can be classified as Devices, as well as Topologies and Components. These model types are grouped under Link.
It is through this modeling intelligence that IMT is capable of inferring the condition of every device on the network, enabling the VNM to isolate faults to the device level by understanding the interaction of network components. When a fault is detected by the VNM, the Artificial Intelligence engine is able to infer that the downstream devices cannot be polled due to a failure of the upstream device. SPECTRUM intelligently filters and suppresses alarms from devices downstream of the fault, only passing through an alarm for the faulty device.
Distributed Management
Distributed SpectroSERVER (DSS) is a modeling feature that uses the concept of landscapes. A landscape is composed of the models, associations, attribute values, alarms, events, and statistics specific to any one SpectroSERVER. Using DSS, networks can be logically partitioned into subnets or multiple landscapes, each with its own local SpectroSERVER. Landscape icons represent distributed server databases, and provide a rollup of alarm information for the devices modeled in those remote databases. Each remote landscape is identified by a unique landscape handle. DSS can improve SPECTRUM performance when managing a large network by distributing the load introduced by management traffic and delegating network management functions to remote workstations.
When a network is modeled with multiple landscapes, it is possible for a single SpectroGRAPH (or other client application) to access information from more than one SpectroSERVER. SPECTRUM's Distributed Data Manager provides access to data from all the SpectroSERVER databases in the managed environment. In addition, applications are distributed. Rather than requiring an application on each server, a single instance of an application can be used across multiple domains, enabling consolidation of alarm monitoring, reporting, inventory, and other key management tasks across the entire enterprise from a single console.
Conversely, the use of multiple SpectroGRAPHs enables division of the IT environment into management domains based on geography, job function, or device type. It also enables an IT organization to distribute administrative responsibility or centralize it in a "control center."
The diagram [click here to see diagram] illustrates SPECTRUM's client/server architecture, as well as device and systems management components of an enterprise solution.
Fault Tolerance (Fail-over Capability)
SPECTRUM's solution for redundancy is based on primary and secondary SpectroSERVERs, also called Fault Tolerance. Having more than one SpectroSERVER managing a landscape provides Fault Tolerance. At any time, only one copy of a landscape is active. The predominantly active landscape is designated the primary SpectroSERVER. The inactive copies are secondary or tertiary SpectroSERVERs. If a secondary SpectroSERVER fails, the primary can be set up to register an alarm that notifies the operators that redundancy has been lost. A secondary SpectroSERVER must be dedicated one-to-one with the primary and mirror its device management modules. When a primary SpectroSERVER fails, a secondary SpectroSERVER becomes active and starts managing the network. Any client applications connected to the primary SpectroSERVER are automatically switched to the secondary SpectroSERVER. When the primary comes back, the applications are automatically switched back to the primary, and the secondary becomes inactive.
Fault tolerance features can be divided into three areas:
- Switching is the mechanism that detects the failure and switches connections to use the redundant resource.
- Readiness measures how quickly the switch to available resources occurs, which is determined by the VNM. A "hot" standby is immediately available when there is a failure. A "warm" standby is running, but may take a short time to be available. A "cold" standby is started when there is a total failure.
- Data synchronization is the measure of how closely the secondary resource tracks changes to the primary resource. It can be measured by the time delay between the change to the primary and the corresponding change to the secondary. It can also be measured by how accurately the secondary server reflects the primary server, which may vary for different kinds of data. In SPECTRUM the data is synchronized by VNM and database mechanisms. One of the database mechanisms is On-Line Backup, which, if configured to backup the primary server, will automatically synchronize with the secondary server.
- The MIB Browser is a graphical interface used to query any SNMP network device and examine its supported MIBs. The MIB Details application is used to call up descriptive information for any monitored object queried from a device.
- The MIB Editor is used to import third party vendor MIBs. The Device Browser tool is used to maintain a database of device IP addresses that are frequently contacted.
ADVANCED APPLICATIONS
SPECTRUM Frame Relay Management Application OverviewFrame Relay is rapidly growing in popularity as an economical and efficient method of connecting offices, especially linking remote offices to corporate headquarters. The benefits of the Frame Relay protocol include lower cost than dedicated T1 lines and multiple connections on a single interface, which allows companies to save a considerable amount of money on routing hardware. As the Frame Relay market continues to grow, so does the need to have real-time information and reports available for each Permanent Virtual Circuit (PVC) in a network. Cabletron offers a scalable Frame Relay management application, fully-integrated with SPECTRUM, that is designed for the Internet Service Providers (ISPs) and corporate user markets. SPECTRUM Frame Relay Manager
SPECTRUM Frame Relay Manager allows Frame Relay carriers, their customers and corporate users to enjoy a greater level of information than has been available from the industry providers in the past. It gives users precise monitoring of Frame Relay circuit bandwidth usage and costs with the use of its detailed reports on bandwidth utilization, line congestion and performance monitoring. Users can easily customize reports to show real-time information at specified time intervals. Features
- Supports all vendors' equipment that use the standard Frame Relay MIB (RFC 1315), thereby eliminating the need for intelligent, expensive, vendor-specific CSU/DSUs, RMON probes or additional management modules.
- Users can manage down to the circuit level, and can monitor Permanent Virtual Circuits (PVCs) of particular interest.
- Provides statistics-per-PVC at user-defined intervals that can be presented in customizable reports.
- Users are notified immediately when PVCs are experiencing congestion or failure, or exceeding user-defined throughput.
- Alarms are generated when load and congestion thresholds are exceeded, as well as in response to a change of state of a Data Link Connection Management Interface (DLCMI).
- Users are immediately notified of changes on Frame Relay equipment.
- The Performance View graphs rest-time throughput and congestion statistics for a PVC.
- An enhanced AutoDiscovery feature allows users to save time while modeling the Frame Relay network, and resolves connections on the DLCMI-to-DLCMI level.
SPECTRUM Frame Relay Manager can monitor any Frame Relay Access Device (FRAD) that supports the standard IETF Frame Relay MIB (RFC 1315) for SNMP data collection. Frame Relay Manager automatically creates an application model representing management support of Frame Relay firmware. Management support of Frame Relay firmware is defined by the RFC 1315 MIB. This MIB contains three groups: one defines the Data Link Connection Management Interface (DLCMI), one defines the circuits, and one describes errors. In addition, one trap is defined which indicates changes to a virtual circuit. The DLCMI table is used to specify various monitoring parameters, such as polling interval and error threshold, and to set the maximum number of virtual circuits. The Circuit table provides information about each virtual circuit for statistics reporting and performance monitoring. The Error table records error information for each virtual interface. The three tables of this MIB can be displayed from this application model. AutoDiscovery
SPECTRUM Frame Relay Manager provides enhanced AutoDiscovery. This maps the Frame Relay network and resolves connections on the DLCMI-to-DLCMI level. A new model type - FrameRelayDTE - visible in the topology view, shows connections of active Frame Relay interfaces, and contains a set of PVCPort models to represent each of the PVC connections. A PVC model is used to indicate that a pipe represents a Frame Relay connection. Alarms
SPECTRUM alarms are generated and users are notified immediately when load and congestion thresholds are exceeded, or in the case of circuit failure. SPECTRUM monitors each PVC for congestion and allows users to define thresholds for throughput. SPECTRUM's fault isolation will identify a down PVC even if the remote router is not modeled. Alarms are generated in response to the MIB RFC 1315 trap that indicates changes in a DLCMI and its current state. Users are notified of status changes before the next poll. Reporting
SPECTRUM Frame Relay Manager provides statistics per PVC at user-defined intervals that are presented in customizable reports, eliminating dependency on reports provided by carriers. These reports can be used to determine a need for greater bandwidth, or to identify circuits with CIR greater than current need.
INTEGRATION
SPECTRUM Partner Integration Partner applications can be integrated with SPECTRUM at three incremental levels: GUI Integration: This level launches third-party applications from either a high level menu pick or an object specific menu pick within SPECTRUM. It can also pass information about the selected object to the invoked application, thus providing a context-sensitive function Event and Alarm Level Integration: At this level of integration, third-party application alarms and events can be viewed from SpectroGRAPH consoles, and automated corrective actions invoked in response. Some partners implement bi-directional alarm/event integration, enabling users to see alarms from both products in either console. Integration at this level takes advantage of many SPECTRUM features, including:- Graphical representation of alarms for the entire enterprise, versus individual managed domains.
- Policy based alarm filtering using SPECTRUM Alarm Notification Manager (SANM).
- Integration with third-party trouble ticket applications.
- Intelligent alarm suppression and correlation
- Integration with SPECTRUM's Enterprise Configuration Manager and SpectroWATCH
- Integrated storage of SPECTRUM and third-party object attributes and statistical data, which can be accessed by SPECTRUM reports, and exported to industry standard databases for further analysis.
- Help Desk/Trouble Tickets
- Capacity Planning
- Facilities Management
- Security/Systems Management
- Data Analysis and Presentation
- Configuration Management
- Applications Management
- Accounting
- Performance Management
METRIX WINWATCH
Today, Metrix's leading product for desktop administration, software distribution and broad desktop management is WinWatch. WinWatch currently supports Windows desktop and servers environments, and will soon support Unix. WinWatch is a highly scalable application that enables flexible, policy-based software distribution, desktop (client/server) configuration, inventory and asset management, remote control/help desk support, and problem management across desktop environments. It is distinguished by its ease of use and ease of deployment. Much of WinWatch's power comes from its intelligent agent technology that provides proactive and efficient, localized management without taxing the network. WinWatch is deeply integrated with SPECTRUM Enterprise Manager and exploits the full breadth of SPECTRUM's knowledge base. WinWatch Architecture The WinWatch PC management system consists of two parts: The WinWatch agents installed on the PC systems and the WinWatch management module, which resides on the SPECTRUM Enterprise management system. The SPECTRUM system is also made up of two main parts: The SpectroSERVER and the SpectroGRAPH. The WinWatch management module consequently consists of two modules: One for installation on the SpectroSERVER and one for the SpectroGRAPH. The following graphic illustrates this basic architecture: [Click here to see graphic] Depending on the size and topology of the network multiple SpectroGRAPHs and SpectroSERVERs may be needed. Below we zoom in on each of these elements to learn more about what they do and how they function. The WinWatch PC System Management Agents WinWatch is a pure Windows program. It does not use any TSRs or special device drivers. This means that WinWatch does not use any conventional memory below 1 MB on DOS based Windows platforms such as Windows 3.1. Being a Windows program implies, of course, that it can only be used on a PC, which runs one of the Microsoft Windows operating systems. WinWatch is not a part of a protocol stack nor does it include any networking software. It is based on the Windows Sockets API, which is an industry standard interface between Windows applications and TCP/IP protocol stacks. This means that to run WinWatch on a PC, that machine must already have some TCP/IP networking software installed and operational. In addition, the protocol stack must support the Windows Sockets API, which almost all vendors do. The MIB supported by WinWatch is a custom MIB defined by Metrix as a superset of the IETF Host Resources MIB. The Host Resources MIB was found to be excellent as the common denominator across multiple platforms, but it was found to be lacking when it came to the most common platform, namely PCs running MS-Windows. The WinWatch MIB addresses this issue of PC management. [Click here to see related graphic] The architecture of WinWatch is based on the idea of extension DLLs providing all actual MIB objects support. The basic WinWatch program (the WinWatch Core) is essentially an SNMP parse engine together with a set of library routines used for common operations. The actual MIB objects supported by the agent, as seen from the network side, are provided by WinWatch Extension DLLs which are loaded by the Core at startup time. Each of these DLLs supports one part of the overall MIB and the operating system functions to which it corresponds. The DLL, using a defined API, informs the Core of the supported objects when it starts. During normal operation, any incoming SNMP requests are handled by the Core, which then accesses the relevant DLL to perform the required Get or Set operation on one or more objects. This architecture, and the defined API, means that the WinWatch agent could be easily programmed to support any MIB, given that one or more DLLs are supplied to provide the support. The WinWatch Management Module The WinWatch management module is very tightly integrated with the Spectrum Enterprise management system, extending its functionality to include comprehensive management of PC systems. The SPECTRUM Enterprise management system is a leading Network and Systems Management platform in the industry, which makes it an ideal candidate for unified IT Management. The SPECTRUM System has a client-server architecture: The servers are called SpectroSERVERs and the clients are called SpectroGRAPHs. In addition there is a comprehensive set of client applications including:- Alarm Manager, which offers global alarm and event monitoring as well as management, and
- Report Generator, which defines, stores and generates reports comprising attributes and statistics from all modeled system and network components.
- New model type objects and inference handlers are added to the SpectroSERVER. The new model types enable full modelation of PC systems, software components as well as the creation of logical PC group objects.
- New views are added to the SpectroGRAPH client. These views provide access to and manageability of many hundreds of attributes on the PC systems.
- The WinWatch Control Panel is a new, `light' client designed for enabling PC systems management on a one-to-one basis.
- The WinWatch Console is a powerful new client designed to support the full set of PC systems management activities.
- Unauthorized new software may be removed and accidentally deleted software can be re-installed.
- Configuration files can be tracked, stored and recovered if modified.
- Full disks may be cleaned of temporary files.
- NT event log entries can scan and initiate corrective actions, etc.
Summary of Features:
Standard browser support The Metrix Web Console is accessible via any browser such as Netscape or Internet Explorer and offers network and system administrators an easy to use interface to manage the enterprise infrastructure from anywhere at anytime. Support for multiple languages The Metrix Web Console supports multiple languages allowing every administrator or operator to access SPECTRUM information in their preferred language. Current languages supported are:- American English
- British English
- French
- German
- Italian
- Spanish
Architecture
The Metrix Web Console is a browser interface with which you can access and manage your network from any computer that has Internet access and a browser installed. The Metrix Web Console is the link between the browser and the Spectrum knowledge base implemented in the SpectroSERVER. It consists of several programs, the most important of which is genview. This is a program based on CGI and takes care of all interaction between the Web Server, the SpectroSERVER and produces the HTML files to be displayed on the browser screen. This following graphic shows the flow of operations through a web server and how HTML pages are created 'on the fly' when a user of a web browser requests information by clicking on a link. [Click here to see related graphic]- The user at the web browser clicks on a link, requesting information. The interaction between the browser and the web server takes place in HTTP.
- The web server calls the custom web application, communicating the requested parameters to it via a CGI-based interface.
- The web application executes the parameters supplied through the requested link and queries the database for the relevant information. The web application and the database communicate in a database specific API.
- An HTML page is created 'on-the-fly' to include the results of the query.
- The HTML page is returned to the web server and forwarded to the browser for display.
- The Metrix Web Console and CGI Interaction
- The browser contacts the server with the requested URL.
- The server acknowledges the request, notices that it points to a CGI script and runs the script.
- The script executes some sequence based on the input from the browser.
- The script generates output formatted in such a way that the Web Server can understand.
- The Web server receives the script output and forwards it to the browser making the original URL request.
- The browser formats and displays the information it received back from the server to the user.
- The browser contacts the Web Server with the URL pointing to 'genview', the main program of the Metrix Web Console, which is based on CGI.
- The server acknowledges the request, notices that it points to the genview program and calls it, passing it the parameters supplied within the URL.
- Genview passes the request, communicates via SSAPI with the SpectroSERVER to find out which view/model it is supposed to display.
- After having found the view/model, genview reads the corresponding HTML file provided as part of the Metrix Web Console and collects all the information needed for this page from the SpectroSERVER.
- After all the information is complete, genview passes the processed HTML page to the Web Server in a format that it can understand.
- The Web Server transfers the output to the browser that interprets it and displays it on the screen for the user.
Read the Patchwork RFP Vendor responses:
