HP shines in test of network management framework tools

Tests show framework-based network-management tools do not have to be complex. Many networks find the framework plus a few modules may be what they need.

A framework-based network-management system doesn't necessarily have to consist of scores of modules that support a supercomplex, hypereclectic computing and network environment. For lots of networks (or if you want to get your feet wet), the framework plus a few modules may be all you need. We call these tools framework express, or framework lite.


How we did itArchive of Network World testsSubscribe to the Network Product Test Results newsletter


Getting to the core of network management by choosing only those modules that support your key devices, servers and applications can be an effective, affordable, productive and smart approach to using a framework-based NMS.

The ideal framework-express package includes a central management and monitoring piece, to which you add a few modules that recognize and manage specific devices, servers and applications. Each module blends seamlessly into the overall NMS, has a small footprint and is easy to use. Modules work together to manage everything, automate administrator tasks, process SNMP alerts (traps), discover the network and diagnose outages and performance problems. The perfect package offers useful reports, scales well, is pervasively platform-neutral and enforces good security.

Network Node Manager traitsTo find a system that meets our criteria, we invited vendors to submit systems to our Alabama lab for testing. We tested HP's OpenView Network Node Manager 7.5, OpenView Operations 7.5 and OpenView Internet Services 6.0; BMC Software's Performance Manager Console 7.5.20, Distribution Server 7.1.21 and Performance Manager Portal 1.2.00; and PerformanceIT's ProIT IT Operations Management Software 4.0. Computer Associates, which had just acquired Aprisma at the time of our tests, said it needed to think about the positioning of Unicenter vs. Aprisma's products and declined our invitation. IBM's Tivoli division, after initially agreeing to participate, backed out of the tests and said a suitable product wouldn't be ready until June.

We awarded HP the Clear Choice Award for OpenView's excellent network discovery, root-cause problem analysis, task automation, responsive and intuitive user interface, and scalability.

System overview

The core of the OpenView framework express is Network Node Manager. In our tests, it excelled at network discovery, device status tracking, network map graphing, statistics gathering and SNMP alert processing. Network Node Manager uses Management Information Base (MIB) data from several sources, including routers, switches, bridges and repeaters. It captures some Layer 2 data, but for the most part it maps Layer 3 details. HP supplies numerous predefined MIB expressions, which Network Node Manager applies. The impressive list includes utilization and error percentages, total packets by category (in, out and errors), retransmits, Cisco memory utilization and full-duplex utilization percentage.

Network Node Manager collects network health data, stores it in a relational database (provided by HP), analyzes the stored device-status and event data, and reports results in useful charts and graphs. The system's root-cause problem analysis, dubbed Advanced Intelligent Diagnosis for Networks, was especially helpful in zeroing in on a specific device that was causing an outage or performance problem, while its path-analysis capability is similarly helpful in pinpointing problems and performance degradations involving network pathways and linkages.

Network Node Manager's quick and accurate discovery feature worked well in all our tests, no matter what mix of devices we asked it to manage. It identified and inventoried not only physical devices, but also virtual network services. Network Node Manager, which accepts what HP terms Smart Plug-ins in order to support new technologies and services, is itself a sort of framework environment.

Network Node Manager's automatic baseline feature, like its discovery feature, makes setup and initial use a breeze. This feature automatically sets alarm thresholds by reviewing and analyzing collected device-status and event data to identify deviations, exceptions and other unusual activity. When we used this feature and added a few thresholds of our own (based on our knowledge of the applications using the network), Network Node Manager thereafter generated prompt and highly informational alarms, via pager or e-mail, to alert us when the thresholds were exceeded.

The system has comprehensive protocol support for packet formats, including HSRP, IPv6 and virtual LAN material. We also found that its distributed architecture scales well to handle larger and more complex network environments. Network Node Manager even monitors itself to ensure it's running normally. It pages an administrator or sends e-mail alerts if the self-monitor finds that Network Node Manager, or its server, has died.

Network Node Manager topology map.

Network Node Manager's extended topology map uses color to show device status and network health.

The OpenView Operations module works with Network Node Manager to provide event management, performance monitoring and automated alert processing. This is especially useful for data centers that need to achieve 24/7 uptime and availability.

We noticed in testing that multiple, concurrently running instances of OpenView Operations coordinated and synchronized with each other, exchanging device and network status information. Running in a clustered environment, OpenView Operations will robustly fail over to another healthy server. HP supplies a rather elaborate programming interface for OpenView Operations, and it sports a high-level Visual Basic Script-like language for customers who want to tailor its processing.

The OpenView Internet Services module excelled at tracking Web transaction-oriented SLA violations. For services we defined, from general Web access to particular e-commerce transactions, it noted availability and response-time details, and alerted us when SLA parameters were exceeded. Alerts took the form of pager calls, e-mail notices and SNMP traps, and we could tell the module to execute a command in response to an alert.

As is true for virtually all of the OpenView modules, each one we tested runs on HP-UX, Sun Solaris, Microsoft Windows (2000, 2003 and XP) and Red Hat Linux.

No agents here

BMC's Performance Manager Portal is an agentless tool that monitors Web-based transactions, as well as IP-based devices and computers. Its three-tier architecture includes application server, database server and Web server components. The Performance Manager console includes nine components: Central Operator Windows, Central Operator Web, Console Server, RTServer, Infrastructure Monitor, Configuration Manager, Console for Unix, Console for Windows and Migration Tools. The Distribution Server remotely installs or uninstalls Patrol components across multiple systems from a single, centralized console. All modules are part of BMC's Business Service Management strategy, which targets applications and their associated network infrastructures.

Together, the Performance Manager Portal and Performance Manager Console identify, diagnose and report simple and complex network problems related to specific application environments, such as Apache Web Server, Compaq Insight Manager, Dell OpenManage, BEA WebLogic, IBM WebSphere, JBoss and SAP. In addition to monitoring specific applications, the system tracks operating-system behavior, including performance, resource consumption and server capacity. It also kept an eye on Active Directory components and events, combed through Windows event logs for items that could trigger alerts based on event type, source, event ID, user or category. The system could also baby-sit individual processes and services to ensure they were running, and noted the consumption of resources. SNMP alerts (traps) could be processed to track network events and errors.

Although the system works without agents, Performance Manager Portal comes with those that you can optionally use. In our tests, agents could provide more information about each monitored system than the agentless environment provided, such as identifying runaway processes inside servers.

The Performance Manager Portal, Performance Manager Console and Distribution Server run on Red Hat Linux, Sun Solaris, Win 2000 and Win 2003.

Adding packs to the system

The ProIT Operations Management Software is similar to monitoring products such as Argent's Guardian and Microsoft's Operation Manager. Yet the base product is still a framework to which you add Management Packs. Each Management Pack monitors a specific application or platform.

ProIT's discovery feature is its AutoMap Dependency engine. Like Network Node Manager's discovery feature, ProIT's AutoMap quickly and accurately found and identified our network components, and it noted network paths and device dependencies as it searched.

ProIT includes three primary components: Infrastructure Services Management (ISM), Applications Services Management (ASM) and Business Service Management (BSM). ISM contains ProIT's core management and monitoring processes, including AutoMap discovery, a notification engine, monitoring engine, a knowledge base for help desk support, operational workflow support and basic reports that show performance and utilization statistics. ASM focuses on specific applications, and is the main interface point for ProIT's Management Packs. ASM also maps dependencies, tracks assets and prepares business availability reports.

BMC Performance Manager product shot

BMC Performance Manager uses an object-oriented approach to showing network elements and events.

BSM contains dashboard displays that show the integration of business and management information. It also supplies reports about application detail, service-level management and business vs. network integration, as well as consolidated system views.

Like OpenView, ProIT has a self-monitor that ensures the system is running. If you subscribe to PerformanceIT's optional external monitoring service, the vendor will remotely monitor each of your ProIT instances from its network operations center.

Each ProIT Management Pack monitors an application, operating system or device, and it comes preconfigured with thoughtfully selected thresholds out of the box. We especially liked being able to apply a Management Pack and its thresholds simultaneously and consistently to an entire group of devices or servers, without having to configure each one. ProIT has Management Packs for various operating systems, including Windows, HP-UX, Solaris, AIX, Digital Unix, Red Hat Linux, SuSE Linux, Debian Linux, Novell NetWare and OS/400. Device support includes those from Cisco, HP, Foundry Networks, 3Com, Extreme Networks, Nortel, Enterasys, Alcatel, Lucent and Juniper. The ProIT Management Packs also support specific applications, including Microsoft Exchange, SQL Server and Oracle databases. ProIT runs on Windows 2000 Server and Windows 2003 Server.

How easy are these things?

HP calls Network Node Manager's user interface Home Base. It gives administrators, engineers, troubleshooters and planners an intuitive, easy-to-navigate summary of the network's status, quick access to detailed alarms and easy-to-understand graphical maps of the network infrastructure and services. HP bundles both a native Windows version and Web-based version of Home Base with Network Node Manager.

Drilling down through Home Base's maps to find device and connection details was a snap - the network health information displayed by the maps helped us locate problems quickly and productively. When a problem developed, Home Base filtered and correlated its store of network events, and produced a summary alarm that described the problem clearly and in plain language. As we investigated a problem, Home Base created and displayed dynamic menus that directed our thoughts and efforts toward solving the problem. Customizing Home Base's analysis of events via the Network Node Manager Correlation Composer feature was simple and straightforward.

Predesigned reports from HP highlight items such as performance, alarm, availability and inventory trends. Many reports contrasted current and historical data, which helped us spot emerging problems, while other reports showed network utilization, top talkers and listeners, and inbound and outbound errors. A Ping Response Time and Ping Retry report showed us such response times and the number of retries, to help measure latency across our network. The RMON Segment Utilization report revealed network bandwidth usage, and a Frame Relay report tracked forward and backward congestion rates to show bottlenecks. Reports also showed summary and detailed device availability, device inventory data, alarm histories and multiple-device reboot events.

The OpenView Internet Services module sported a productive dashboard interface, offering us a quick, tree-based navigation, SLA health indicators and a helpful troubleshooting and analysis tool.

Business-eye view

BMC's Performance Manager Portal is a Web-based console that displayed views of the network infrastructure elements, as well as views of the business applications corresponding to those elements. BMC designed the dual network and business views to help customers more quickly identify and fix problems. Although switching between views was simple enough, we found the dual views didn't appreciably aid us in determining the problem. When a router fails, for example, its relationship to the business environment isn't as important as deciding what's causing the failure and knowing how to solve it. However, Performance Manager Portal gave us a level of business intelligence, as it clearly showed us the status and health of the infrastructure elements and how those elements contributed to our overall business environment. We also felt that Performance Manager's user interface wasn't as responsive as the Home Base interface.

The Performance Manager Portal includes console interfaces for administering, operating, configuring and distributing the Performance Manager modules. Via the Portal, we could see the network, servers and applications as objects displayed in an expandable and collapsible Object Tree. Clicking on objects drilled down to details regarding the health of applications, services, servers and devices. The Portal and Performance Manager Console module provided us with the same breadth of reports as HP's Home Base.

ProIT's Web-based user interface gave us four types of network maps: X-type, in which device icons are logically organized into an X shape with gateway devices at the center; concentric, in which devices are drawn in concentric circles radiating from the gateway at the center; ring, in which devices are drawn in a single circle; and table, in which devices are drawn in rows and columns. ProIT showed the direction of dependencies along network paths, and it differentiated between direct dependencies and secondary ones by drawing direct dependencies in bold, and secondary ones in gray. Next to the map, it displayed devices without clear dependencies, letting us investigate why the device appeared not to have any dependencies.

OpenView Network Node Manager 7.5, OpenView Operations 7.5, OpenView Internet Services 6.0OVERALL RATING
4.5
Company: HP. Cost: Network Node Manager, $6,000; OpenView Operation, $17,995; OpenView Internet Services, starts at $12,449 for 5 targets. Additional targets: 5, $2,038; 25, $10,207; 250, $65,160. Pros: Handles large, diverse networks with ease; intuitive interface; useful reports. Con: No printed documentation.
Performance Manager Console 7.5.20, Distribution Server 7.1.21, Performance Manager Portal 1.2.00OVERALL RATING
3.6
Company: BMC Software. Cost: Performance Manager Portal, $4,000; Performance Manager Console, $525 per CPU; Distribution Server is included in package. Pros: Excellent discover; good reports. Cons: Sluggish interface.
ProIT Operations Management Software 4.0OVERALL RATING
3.6
Company: PerformanceIT. Cost: Starts at $9,995 for Standard Edition (50 devices). Pro: Good monitoring of applications and services. Con: Runs only on Windows.
The breakdown  HPBMCPerformance IT

Management 20%

534
Reporting 20%543
Ease of use 20%533
Corrective action 10%444
Notification 10%444
Installation 10%444
Documentation 10%344
TOTAL SCORE4.53.63.6
Scoring Key: 5: Exceptional; 4: Very good; 3: Average; 2: Below average; 1: Consistently subpar

Like the HP and BMC systems, ProIT showed us network health and device status information on its maps. Clicking a map element drilled down to detailed statistics and status information. The user interface was more responsive than Performance Manager's, but not as responsive as the Home Base interface from HP.

ProIT includes several preconfigured management reports, as well as SLA management, availability management, capacity planning and trend-analysis reports. The SLA reports were particularly well designed, showing us the current and previous period's (usually a month) statistics for historical trend assessment. Other reports showed us application, server and network uptime statistics, profile information for asset tracking, event-log analysis, alarm histories and utilization data.

Putting it together

BMC and PerformanceIT gave us printed and online documentation with their systems, while HP OpenView documentation is only online. Despite their complexity, all products were easy to install and begin using.

Overall we felt the HP Network Node Manager, OpenView Operations and OpenView Internet Service modules were collectively an excellent way to explore framework-based network management and monitoring. The system will excel for a growing midsize company that needs scalable tools with greater capacity and more functions, or for a large company that wants to manage and monitor its network more closely.

Nance runs Network Testing Labs and is the author of Introduction to Networking, 4th edition and Client/Server LAN Programming. He can be reached at barryn@erols.com.

NW Lab Alliance

Nance is also a member of the Network World Lab Alliance, a cooperative of the premier reviewers in the network industry, each bringing to bear years of practical experience on every review. For more Lab Alliance information, including what it takes to become a member, go to www.networkworld.com/alliance.

Learn more about this topic

IBM offers up best practices for software development

10/12/05

Start-up debuts with an all-in-one security platform

06/27/05

AT&T collaborates with Microsoft

06/13/05

1 2 3 4 5 Page 1
Must read: 10 new UI features coming to Windows 10