Printer Friendly

Network Management Adjusts to the New Era.

Fueled by user needs for enhanced capabilities and greater operational simplicity, network mangement systems are adding more and more automated functions and support for heterogeneous network configurations incorporating equipment for many vendors and a mix of analog, digital and private-wire facilities. Automatic generation and tracking of trouble tickets is common with today's systems, and some even offer automatic fault analysis. As for operational simplicity, the latest systems use color graphics terminals with icons representing network functions, supplemented with "help" screens and other operator prompts. Besides supplying more detailed, complete and up-to-the-minute information on network operations, the systems alsl collect network performance data so that the mangaer can optimize the network.

Typically, a network-control system employs a master controller at the central site, which communicates with a special electronics module added to each modem in the network. This module is resonsible for gathering status and operational data from its associated modem and responding to commands from the controller. Communications takes place over the regular data network, but in a low-speed band separate from the main channel. This means that diagnostic and control commands and responses can flow between the controller and the remote modules with interrupting regular network operations. At the central site, a CRT terminal alerts the operator to network problems and provides the means for diagnosing a problem, isolating the faulty unit and restoring network operation while the failed element is being repaired or replaced.

For most of the time, network-control systems operate in an automatic monitoring mode, where the controller polls each modem in turn, alerting the operator when any condition exceeds pre-set threshold limits. Diagnostic tests can then be run to identify the nature taken, problem, and corrective action taken, such as initiating dial backup, switching to a hot-standby modem, or disabling a streaming modem.

IBM takes a completely different tack to network control, rejecting the use of a secondary channel and interspersing the monitoring control signals with the data stream instead. The network management functions are distributed among the various network components of a typical IBM SNA network. In the communications controller, the network-control program assists the network manager in analyzing activities within the communications controller, as well as the lines and devices attached to it. IBM's microprocessor-based modems and the 3867 link diagnostic unit provide information about line problems and line signal quality; the modems also maintain status indicators. SNA controllers collect network-management statistics, such as error data and SDLC link test results, and display them on request to network control center personnel.

At the host site, the Network Communications Control Facility. (NCCF) and the Network Problem Determination Application (NPDA) program products enable the communications manager to exercise the various network-management capabilities distributed throughout the network. In non-SNA networks, users can still run local and remote loopback and similar diagnostic tests with the modems, but they must be done manually from an operator's panel on the modem.

When the network is operating normally, its status is sampled continuously at pre-set intervals, and information on line, modem and terminal interface conditions is routed to a centralized data base. If intermittent problems occur, the network operator can access this data base to check trends in error performance or other parameters that might be causing the trouble. In addition, the operator can initiate tests to validate problems or verify repairs.

To isolate faults, the central-site softwaree operates in conjunction with the Link Problem Determination Aid (LPDA) function that is designed into each modem. Among the diagnostic capabilities provided by LPDA are local and remote modem self-test, and local and remote loopback tests. In addiiton, the operator can request reports on the status of local and remote modems and the remote terminal interface.

At the heart of the network mangaement system is the Network Communications Control Facility (NCCF) program product, which allows the operator to control all or a portion of the network. Several operators can each control different network components or one operator can control a complete network, including multiple domains and multiple applications. NCCF simplifies the development of automated responses to operator messages, thereby reducing operator effort, error potential and skill-level requirements.

AT&T Information Systems' Data-phone II service makes use of a secondary channel and microprocessor-based modems operating at 2.4, 4.8 and 9.6 kb/s to provide four levels of network control to accommodate configurations of different sizes and complexity. Also, the service enables users to upgrade from one level to the next in a virtually transparent manner. With the first level of network control, users can designate a central-site modem as a control unit. This modem continuously polls all tributary modems in accordance with a polling list entered by the user. In addition, the control modem periodically polls a special address to determine if any tributary has requested a maintenance test. If so, the control modem runs the test.

Level II service features a diagnostic console that provides expanded test and command capabilities and allows users to extend the duration of certain tests and to monitor and test any modem in the network. Level III system employ the network controller, a multi-port device with a microprocessor-controlled cartridge tape unit. Mulliple tests and commands can be queued automatically or delayed for execution at a time specified by the user. Test routines and network layout information can be stored on the tape cartridge, which also contains the system instructions.

Dataphone II Level IV, the most advanced product in the Dataphone line, has the ability to manage and control larger, more complex data networks than any similar system, according to AT&T. In February, the firm announced three multiplexers designed to operate with Level IV, giving network managers the opportunity to lower line costs while mataining the diagnostics and control features of their present system. One product, the 718 Star Mux, can handle from four to 32 channels of asynchronous and synchronous data at speeds to 9.6 kb/s on each channel with virtually error-free performance. It offers the flexibility to work with coaxial, fiber-optic, satellite and metallic transmission facilities. Another unit, the 735 T-Mux, digitizes and routes voice and video traffic, as well as data, over the TI lines. The third product, the 719 Networker, is a combination statistical multiplexer and data packet switcher for routing multiple remote locations into a central control. According to Ken White, manager for data communications equipment, the 719 is "the intellgient link to the Dataphone II Level IV." This particular product allows users to reconfigure all multiplexers from a central location and sends all error conditions to the diagnostic system.

"The 719 Networker is a stand-alone 32-channel computer port concentrator for local network use," says White. "Or, you ca use its data packet-switching capabilities to tie stat muxes into a single transmission facility. You can even tie up to 250 Networkers together. For example, multiple bank branches can communicate in this way with a regional center."

More recently, AT&T introduced an access controller to serve as an interface between the Level IV system controler and the firm's Dataphone multiplexers. The new device continuopusly monitors and controls all Dataphone multiplexers under the command of workstation terminals of the system controller. AT&T also enhanced the Level III network controller, extending the controller's monitoring and diagnostic management capabilities to more complex configurations, while cutting peripheral eqauipment costs by up to 40 percent. The savings come from being able to use less-expensive AT&T Model 4425 video display terminals, Model 475 printers and the recently announced 2212C modem. Software enhancements extend the controller's monitoring and diagnostic management capabilities to network configurations with tandem multippoint circuits.

Atlantic Research's Network Test and Management System is a comination of test equipment, patching and switching modules and data-base facility designed to provide fingertip control of three network management functions: tech control, traffic monitoring and administrative management. For each control, the Alexandria, Virginia firm uses an electronic switching facility with manual backup that allows the operator to quickly restore service by bypassing faulty lines and modems and switching to backup circuits and terminals automatically. All switching functions are under software control and there is an array of diagnostic test equipment for pinpointing problems so that faulty lines can be restored rapidly.

Automatic patching combined with sophisticated line monitoring and test sets allows the operator to generate data on processing and response times, traffic loading, port utilization and other statistics vital to network operation. The operator can also perform interactive testing, analyze performance automatically and spot troubles before the circuits deteriorate. As for administrative management, the on-line data-base facility allows the network manager to keep track of the entire inventory of circuits, modems and terminals and maintain service records, cross references and special trouble tickets on faulty circuits.

Atlantic Research recently introduced a CRT-controlled fallback switch system that allows the user to reconfigure up to 1500 circuits per site under both local and remote control. According to the firm, the system offers a cost-effective step between its manual patch/switch systems and fully automated network test systems. At its heart is a new entry-level controller that can later be upgraded to an automated network test system with advanced alarm, trouble ticket data base and test access capabilities without replacing existing equipment.

Avant-Garde Computing offers a number of network management tools, including the Net/Alert network performance monitoring and management system, the Net/Link network planner and analyzer, the Net/Switch electronic matrix switch, the Net/Measure performance monitor for small networks and the Tempo response-time monitor for individual terminals. In March, the Cherry Hill, New Jersey firm introduced a software-based system, Net/Command, which simplifies the management of large data networks with multiple host computers and sites by consolidating many network control and management tasks in a single system. Net/Command provides access to, and control of, multiple products and functions. It provides vendor-independent centralized control from a single workstation of the many tools used for network status and performance monitoring, problem identification and diagnosis and network reconfiguration.

Avant-Garde has also introduced a system to control and monitor PC access to data networks. Called Net Guard, the stand-alone system enables connection, control and monitoring of personal computers and other dial-up devices accessing data networks. It also allows communications with dial-up users from a network control center. Net/Guard connects any dial-up device to the network, regardless of protocol, and controls access to the system data base and applications through the use of multi-level passwords. The system monitors, measures and analyzes current and historical status, performance and utilization of the dial-up devices, providing displays, charts and reports via color monitor and printer.

With Net/Guard, a network manager can see who is using the network and for what applications, providing greater end-user service and credibility without losing central control. Besides protecting network and data base integrity against unauthorized access, Net/Guard also identifies potential functional and performance problems and allows managers to evaluate service levels and plan for growth. Net/Guard provides three types of alerts to network operators: security alerts occur when an invalid log-on is attempted; equipment alerts identify hardware malfunctions or failures; and threshold alerts occur when a dial-in user exceeds a specified connect or idle time limit. An audit trail provides a history of log-ons, alerts and other system activity.

Case Rixon Takes Unified Approach

Case Rixon launched its network-management strategy in March with the announcement of its first network-management processor, the 5200, a 32-bit computer with color graphics workstations. According to the Silver Spring, Maryland firm, all Case networking products, both current and future, will be provided with appropriate "probes" that monitor network performance and pass information back to the controller for processing. Examples of such probes include modem diagnostic cards, wraparound units for nondiagnostic modems, and network-control-and-management modules for Case DCX multiplexer and switching products. In addition, a common interface on products under development will allow central-site management of data service units, channel service units, T1-type multiplexers, local-area networks and new generations of networking products.

The Case 5200 manages networks of 120 to 720 lines, supporting up to 9,000 devices. It embraces all network-management functions, from the highly interactive elements of operational control to the compilation of statistics for long-term network development. At the operational level, system faults trigger alarms, and trouble tickets can be constructed for the implementation of remedial action. Sophisticated switching devices implement re-routing procedures locally, reporting the results back to the controller. In the case of modem failure, the 5200 restores transmission through the switched telephone network. For administrative purposes, data traffic details provide the accounting information for internal billing, while loading and performance logs enable the planner to extend the network as business develops.

To ensure easy transfer to other processors as larger management systems are required, Case is developing the software under the Unix operating system. Unix also allows inter-processor communications when multiprocess configurations are chosen and supports the Graphic Kernel System standard as the base for graphical representation.

At the ICA show in Dallas, Case added to its 5000 Series of network-management systems by unveiling the 5100 system for small and medium-size networks. The 5100 manages networks of 30 to 120 lines, supporting up to 1500 devices. It employs a 32-bit central-site controller, which supports up to four color workstations. Other features include ticket management, multilevel operator security and a sophisticated relational data base with a 26-megabyte hard disk and 640K floppy disk.

Codex's latest network-management-and-control system, the 4800 Series, addresses the large-scale network requirements of Fortune 500 users whose communications networks represent a critical corporation asset. The Series comprises two basic systems: the 4860 can monitor the information flow over as many as 744 lines; the 4850 is a smaller version of the 4860 and can handle up to 496 lines. Both models come with a 32-bit minicomputer and microprocessor-based Distributed Network Processors, which support simultaneous, continuous polling of the network. Each DNP offers expandable line support up to 124 lines. The Codex 4850 supports up to four DNPs; the 4860 supports up to six.

The minicomputer collects and processes data for the entire network, and provides the power for all network control, monitoring and management functions. The 4850 comes with 50 megabytes of disk storage and two megabytes of main memory, while the 4860 features 73 megabytes of disk storage and four megabytes of main memory. The network control terminal serves as the operator interface to network-control-and-management applications; the 4850 supports up to six NCTs, and the 4860 up to 10.

Among the network-control features are continuous on-line monitoring and the ability to initiate a number of comprehensive analog and digital tests from the NCT, either from a central or remote site. In addition, the network manager can change the modem data rate and reconfigure its application from the central site. The manager can also replace a failing modem with a hot spare from the central site with minimal disruption of service.

Codex Management Applications (CMA) augment the system features with five additional functions for management reporting. The menu-driven software provides network managers with extensive information, statistics and flexible reports on the performance of the entire network and allows managers to identify trends and undertake long-term network planning.

According to the Mansfield, Massachusetts firm, the 4860 was designed in cooperation with Delta Air Lines, one of Codex's key customers. Codex claims the system is the only one of its kind in operation, monitoring and controlling an on-line network of over 3,000 modems. Implementation of the Codex 4860 at Delta also makes the airline carrier's network one of the largest private data networks under centralized control.

At a briefing announcing the system at Delta's Atlanta headquarters, Codex Vice President Murray Bolt described the 4860 developments as a response to the communications requirements of large corporations. "The communications network is emerging as a business tool, indispensable to the profitable and productive operations of a corporation," he says. "The more essential the network becomes to a company's business, the greater the imperative to overlay effective network management and control.

Bolt also claims that few other communications vendors are tackling the problem of network control and management in the "heterogeneous" network environment. "The heterogeneous, or multi-vendor network, is definitely emerging as the predominant environment, reflecting the user's growing insistence on choices--the selection of data processing devices based on capability rather than just compatibility," he says. "In the heterogeneous environment, the communications network emerges as the 'facilitator' among all these different computing devices--and the problem of centralized management and control becomes one of the biggest challenges."

The work done on the 4850 and 4860 is part of an "architected approach" to network management and control under continuing development at Codex. According to Bolt, the firm has contributed to the formulation of a set of proposed architectures and standards that will allow management and control of heterogeneous networks. These documents will be part of US proposals for adoption by the International Standards Organization (ISO).

Architecture for All Products

What Codex has done is to define the components of network control and management and their interrelationship. From this, the company derived a model, which, in turn, is being translated into an architecture emcompassing all of its products. The architecture can be easily incorporated into other manufacturers' products as well, the firm claims. The architecture distributes certain levels and types of diagnostic information to a set of devices, called agents, that are strategically located in different networks and/or products. The agents then collect, and in some cases process, network data and transmit it back to what Codex calls an applications processor. This unit does another level of processing and passes the information to the operator interface for notification and action. The information can also be collected for future analysis and planning.

To integrate a proprietary network, for example IBM's SNA scheme, into this heterogeneous environment, Codex would simply embed the SNA agent at the application level in the IBM CPU and transmit the architected data back to the applications processor. In essence, this gives the user a single window on the modem-based network, nodal processor network, SNA network, local-area network, and so on, for the purposes of control and management.

"The development and standards work Codex is doing in network management and control will allow the user to increasingly optimize the overall cost/performance of the network, regardless of how complex an infrastructure it may really have," says Bolt. "What we will have achieved is a more comprehensive answer to network management and control in the heterogeneous environment--an approach that is simply not available to the user today."

Comm+ Systems, an unregulated subsidiary of Pacific Northwest Bell, offers a number of software-based products that have evolved from integrated network-management systems developed at Pacific Northwest Bell as early as 1979. For companies with in-house network-control centers, Comm+ offers a Data Network Administrative System (DNAS) that integrates database-management and performance-measurement/reporting capabilities. With DNAS, the network-control center can monitor, isolate, measure and manage complex computer networks with a high degree of automation.

Comm+ also offers an automated modem-independent system, called Jason, for the simplified testing of data-network circuits. Jason concentrates control of circuit testing at a single terminal in a single location. There, an operator executes menu-driven diagnostics and attaches to the network complex test equipment, all with keyboard commands. Smaller companies can also benefit from the sophisticated control features of both the DNAS and Jason systems through an off-site control facility available to companies on a contractual basis. Based in Seattle, the Remote Network Service Center (RNSC) gives customers a single point of contact for reliable problem-solving. RNSC specialists can remotely isolate and diagnose a problem, and either fix it over the phone or dispatch the proper service. RNSC also provides reports on component and vendor performance.

DNAS is priced at $70,000 and requires an IBM 370 or equivalent operating under an OS/MVS or MVS/XA operating system. DANS gives network operators a macro-to-micro view of their entire network configuration with color graphics on one screen. Different colors indicate different network conditions; on command, the graphic will show the next lower level for more monitoring detail. DNAS provides a data base and communications vehicle for tracking the multiple aspects of an installation. It also helps the network operator track the implementation of changes in the network and provides reports which serve as long-range planning tools for network designers.

Jason presents all test readings on the control terminal screen without requiring the network operator to touch the test equipment. Operators can test all elements of a transmission path, including lines, modems and terminals with one call. Tests encompass quick scans and in-depth analyses of the analog transmission paths in the network. Jason can also be programmed to run tests on a regular basis for network monitoring. Jason's software is supported by a Digital 11/73 minicomputer, which is shipped with the system. Including hardware and software license, Jason costs $118,400. Future enhancement plans include test and diagnostic capabilities for digital circuits, T1 carriers and private switched networks.

Comm+ also offers Jason as a service, allowing a user's technicians to remotely test and diagnose their network by dialing into the Jason system at Comm+. The firm also provides a consulting service to companies that need a third-party network-control service, but are looking ahead to where network growth will justify an in-house control center.

PC Used as System Controller

Digilog's new Network Diagnostic and Test System (NDTS) overlays existing networks and uses an IBM PC as the system controller. NDTS accesses RS-232-C, V.35 and VF circuits and can handle 256 lines. From the PC console, the operator can monitor fault alarms and network status, test system circuits and perform A/B switching. NDTS utilizes a data base that identifies circuits and equipment with user-defined names. Entering a circuit name into the PC gives the operator instant information on the circuit and any comments or problems. On detection of an alarm, the system posts the circuit identification, a description of the alarm and a date-time stamp both to the CRT and system printer. Through the CRT console, the operator can then access the circuit for monitoring or testing to determine the exact nature of the problem.

Once the problem has been identified, the operator can then decide to use the NDTS A/B switching feature to restore the circuit to operational status. For ease of use, NDTS also provides a series of "test busses" and a test equipment access matrix that automatically connects the proper test equipment to the circuit under test. The operator can also request a "snapshot" of an alarming circuit. NDTS will display the RS-232-C signal levels or state of activity, the transmit and receive analog levels and the current position of any A/B switches.

Digilog's Network Control Division has also redesigned its Network Access Switch (NAS) to handle more ports, more lines, more types of circuits and higher speeds. When a fault occurs in the network, NAS alerts the operator to the problem, provides the test access necessary to isolate the fault, and enables the user to switch in spare modems, lines, ports or other devices, either at the central site or remote locations. NAS is compatible with all popular codes and protocols at data rates to 56 kb/s. It provides access to RS-232-C, V.35 and four-wire analog circuits. When fully expanded, the system can accommodate up to 1440 analog or digital ports.

The Montgomeryville, Pennsylvania firm also offers a Modem Diagnostic Test System (MDTS) that adds intelligent capabilities to any vendor's modem for automatic digital and VF alarming, modem and terminal status checking, remote testing and "streaming" device detection and correction. The system is controlled by an IBM PC program written by Digilog.

"Using our newly developed 'wrap box' we can equip each modem, regardless of the manufacturer, with the capabilities for central alarming and remote testing," explains Mike Markulec, product director. "Once the modems are wrapped, the user has central console control of the network."

The NDTS central workstation monitors the operation of the network modems and maintains a record of alarm conditions. Users can selectively enable or disable each monitor function, set high and low limits for alarm processing and check the current state of each monitored point. From the controller, the operator can command an individual Wrap Around Box, WRB/1, or a group of them to perform interruptive tests on the network. The Wrap Box interfaces with virtually any four-wire, leased-line modem that operates from 1.2 to 9.6 kb/s. It provides a secondary channel over the same communications lines used for the network.

Dynatech Data Systems' latest network-control system, the DynaNet 240, combines matrix switching with real-time interface monitoring and alarming. If it detects a condition outside the user-defined parameters, the DynaNet 240 alerts the operation staff with an audible alarm and a full-color display of the name of the alarm, the port or group affected and the nature and priority of the failure. The system allows users to define two alarm conditions for each line; each may be different from the parameters defined for any other line. All alarm parameters are monitored with one-millisecond resolution. The appearance and disappearance of each alarm condition is automatically recorded on an audit trail printer and stored in memory for future reference in management reports.

Data Resource Management

The Springfield, Virginia firm also moved into the area of performance measurement with the recent introduction of Prism, which can stand alone as a performance measurement, monitoring and reporting system, or be integrated with the firm's CTM series of matrix switches for complete data resource management. A single system can measure thousands of lines, with each line having hundreds of entities, allowing virtually unlimited expansion.

Protocols can be downloaded to the sensors, enabling the users to easily migrate, for example, from bisync to SDLC, or to move sensors from one line to another. A complete network view is provided by melding information collected at the local site as well as from remote locations. All the Prism capabilities are available from an entry level, of, say, 16 lines, to a fully expanded system comprising thousands of lines. Time-driven procedures, such as weekly performance summaries, can be activated automatically without human intervention. A typical 32-line stand-alone Prism system costs less than $2,000 a line, with delivery scheduled to begin in the fourth quarter.

General DataComm Industries' Netcon network-management system provides non-interfering management beyond analog modem networks; Netcon's control can be extended to multiplexing systems, digital (DDS) networks, wire lines, local-area data sets and financial loops. Netcon provides continuous network surveillance, with real-time alarms when user-selectable thresholds are breached. Netcon components include diagnostic modems and a "wraparound" capability for mixed-vendor non-diagnostic modems. Diagnostic test routines can be predefined and stored, and programmed for a time when operations are impacted the least. A "help screen" provides the operator with specific background information and parameter prompting, alleviating the need for comprehensive operator training. Service-restoration options include automated dial back-up, modem sparing and A/B switching. Netcon also maintains records of historical alarms, trouble tickets, station circuit data and critical statistics on network performance. If need be, the data base can be transferred to a host computer for additional analysis.

System for Small and Medium Networks

The latest addition to the Netcon family, the NDC-40 network controller, is designed and engineered for small to medium-sized networks. The NDC-40 serves up to 64-multipoint lines and 1,024 diagnostic addresses, providing real-time reporting of network alarms in point-to-point and multipoint networks over analog, digital and wireline facilities. The dual-diskette-based controller comes with a color CRT console, accepts an optional printer and supports all current and planned Netcon hardware. According to the Middlebury, Connecticut firm, Netcon's full complement of diagnostic tests is available to the NDC-40 operator at the unit's color CRT terminal. As with all Netcon routines, the operator is guided by user-friendly English language prompt references and "help" files.

Infinet has integrated a new series of network management, network control and performance measurement products into a single data base in response to what it sees as "an industry-wide need for increased network-management flexibility and control over analog, digital and hybrid networks." The menu-driven NIS 90/60 is a custom decision-support system that uses simple "query-by-example" techniques to store and retrieve data. According to the Andover, Massachusetts firm, the system is easily programmable, tracks problems and availability, and speeds problem determination and resolution.

Featuring an open architecture, the 90/60 accepts and processes line-parameter data and protocol-related information from the firm's EMS-II Network Control System and the PMS-II Performance Measurement System. The two monitoring functions provide users with information on the physical and logical network in one data base. The 90/60 features a flexible report generator that automatically assembles and formats data from the data base and presents the information in graphical and tabular format. The great value of the 90/60, according to Jim Capeless, vice president of marketing development, "is the flexibility with which users can create and update reports to complement particular management styles."

Can Automatically Open Trouble Ticket

As an example of its management report flexibility, Capeless cites problem management, a custom trouble-ticket system used to track network problems. Specific ticket formats can be modified to reflect unique equipment situations, he says. The system can be programmed to automatically "open" a trouble ticket when user-preset thresholds for network quality are exceeded. The ticket is continually updated and finally "closed" when the problem is resolved.

The 90/60 utilizes a Digital VAX 11/730 processor running under the Unix operating system and will support 16 terminal devices, including multiple EMS systems. EMS-II allows users to monitor, test and control both digital and analog lines and devices from a single console. The integration of analog and digital diagnostics eliminates the need to distinguish between transmission types when designing a network, Capeless notes.

EMS II provides advanced analog parameter measurements, including phase jitter, signal-to-noise ratio and received-signal level. Analog diagnostics and control extend beyond a hub, reducing the need for remote-site technical personnel. EMS-II also measures line quality the same way the telephone company does by generating a 1,004-Hz tone. These advanced diagnostics speed problem resolution and reduce vendor fingerpointing, Capeless claims.

The EMS-II responds to changing network needs with the ability to modify network devices. Modem speeds and multiplexer channels can be reconfigured from a central-site EMS console. From the same console, users can restore network operations with dial backup for both analog and digital lines. Capeless points out that this is especially important because of time-consuming repairs required for digital lines. EMS-II also allows users to actuate spare or internal fault-tolerant modems and disable streaming terminals or modems. The console is menu-driven and features color-coded function keys for enhanced ease of use.

The PMS-II monitors performance and protocol-regulated parameters such as line utilization, traffic count, error count and response time for planning network growth and change. This is accomplished using intelligent line-monitoring units (LMUs) that monitor 3270 SDLC and bisync protocol-related activity and send the information to the 90/60 data base. According to Capeless, the ability to handle mixed protocols allows the user to implement change management in large dynamic networks. LMUs are passive devices that monitor data without interrupting main-channel data flow. Because LMUs are intelligent, information is pre-processed, off-loading the 90/60 data base for other control and management tasks. PMS-II also warns users when performance deviates from user-preset performance thresholds.

In March, the Sunrise, Florida firm introduced a matrix switch to its CMS line to provide system redundancy for all the major network devices. If a device fails, the CMS Matrix Switch can restore the network to full operating capacity in a matter of seconds by reallocating port assignments from the failed unit to its designated back-up spare. The switch supports all major data communications interfaces, including RS-232-C, V.35, analog and current loop. It is completely protocol-transparent, supporting both synchronous and asynchronous data at speeds of 76.8 kb/s. Up to 16 mainframe units may be configured together, allowing a maximum of 3840 V.24, analog and current loop ports, or 960 V.35 ports. In addition, up to 63 remote matrix switches can be configured for central-site control.

Alarms on an Unsolicited Basis

According to Larry Pigeon, director of marketing for the firm's Communications Network Division, the CMS family is the only one that provides alarms on an unsolicited basis. "No polling is involved," he explains. "Consequently, failure alarms are communicated on a true real-time basis." Once an alarm has been recognized, a full complement of on-line tests can be initiated by the console operator to determine the nature of the problem. "These include, but are not limited to, end-to-end tests, digital/analog loopback tests and power failure tests," Pigeon explains. "After testing to isolate the fault in the remote modem, the operator can call up the matrix switching software menu on the same CRT, and, by means of a simple one-step command, bring a designated spare modem on-line for replacement."

To allow the CMS system to function with other manufacturers' modems, Racal-Milgo offers a "wraparound device," the TCM-7, which can operate in point-to-point or multidrop configurations at rates from 1.2 to 16.8 kb/s. With the CMS system, such as modem operations as fallback rates and strapping adjustments can be managed from the central site. The CMS 2000 series also offers a mux-control interface that allows the operator to establish and change key operating parameters for on-line multiplexers.

One feature of the CMS system, called AP-II, can be used to perform the same type of line testing as that formerly provided by AT&T and now by the regional Bell operating companies (RBOCs). If line problems are suspected, the network manager can use data from analog threshold alarms and other sources to make exact and authoritative line measurements of attenuation distortion, impulse noise, signal-to-noise ratio, phase hits and several other critical parameters. This data can then be relayed to the respective RBOC for corrective action. In the CMS software, all Bell specifications for both conditioned and unconditioned lines are pre-loaded into the software, and presented in a four-color-coded graphic display, so that information obtained can be conveyed to carriers in their own terms for much quicker system restoral.

Another formerly independent function that has been integrated into the CMS system is a response-time analyzer. The optional software, known as Span (for System Performance Analyzer), can be used to generate user-defined threshold alarms for the response times involved in terminal-to-host line transactions. The response times can also be tracked over variable time segments, and entered into a host-transferable data base for future network trend analysis. In addition, Span will monitor and record line utilization.

T-Bar has added the PMS/1060 network-management systems developed by France's Thomson-TITN to its line of switching and control hardware and software. When used in combination with T-Bar's switching products, the PMS/1060 provides users with centralized control of their networks and enables them to analyze performance and to identify and correct problems ranging from poor response time to equipment and facility failure, claims A. Henry Morgan, chairman and chief executive officer of the Wilton, Connecticut firm.

Switching Between Protocols

"In at least three critical areas, the PMS/1060 gives users more flexibility than any other network-management system," Morgan claims. The PMS/1060 is the only on-line system that makes it possible for protocol changes to be downloaded without the need to disconnect probe hardware or change firmwear, he says. This characteristic, Dynamic Downloading, allows switching between protocols by simply entering a software command. Competitive products require unplugging boards and PROMs to change protocols and line probes, requiring their performance measurement systems to go off-line, Morgan states.

Secondly, the PMS-1060 allows for selective or continuous monitoring. Combining the system with T-Bar's Virtual Switch Matrix makes it possible to implement dynamic line monitoring in random or sequential modes. In addition, the multi-processor PMS-1060 allows users to pay for as much processing power as needed and provide for graceful degradation in the event of a failure. Competitive products typically feature a single processor and thus are more sensitive to catastrophic failures, Morgan claims.

Real-time color console displays allow users to generate on-line system profiles from individual or multiple display consoles simultaneously. The PSM/1060 can monitor from 16 to 512 lines with multiple protocol capability, including SNA 3270, bisync, asynchronous and Honeywell VIP. Capability for X.25 packet-switching operation will be available in the third quarter.

Identifying Failures

Multiple levels of "zoom" help users immediately identify system-element failures. Terminal responses, for example, can be verified to see the amount of time it takes for a transaction and what needs to be refined to improve response efficiency. T-Bar also offers an optional data-analysis system, which allows users to keep cumulative records automatically of their system performance and operations. The reports can be generated to spot underutilization of lines or excess equipment capacity. Morgan reports that the PMS/1060 and the DAS/1061 data analysis system will be integrated into T-Bar's Overlord data-system resource manager as its performance measurement and data analysis modules. This will permit the addition of "anticipation" to the threshold alarm signaling that already exists in all T-Bar switching equipment, Morgan states.
COPYRIGHT 1985 Nelson Publishing
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1985 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Edwards, M.
Publication:Communications News
Date:Jul 1, 1985
Previous Article:Change in PBX Marketplace Forces Users and Vendors to Alter Plans.
Next Article:How Network Analysis-Management Software Supports the Telecom Manager-Part II.

Related Articles
RUSSIA - Gazprom.
Viewpoint--"computing on demand".
Pay for performance--for whom the bell tolls.
Did you know ...

Terms of use | Copyright © 2016 Farlex, Inc. | Feedback | For webmasters