Components of WBM | Web-Based Management

Components for WBM can be written in any language, including HyperText Markup Language (HTML), Common Gateway Interface (CGI) scripts, and Java.

HTML provides a standardized way to create pages of Web-based information and embed hyperlinks that lead the user to other pages. While HTML pages are usually static and textual, they can be made interactive by embedding graphics and active elements, such as Java applets and CGI applications, within the page. HTML is the ideal tool for displaying tables, such as network inventory details and IP address listings.

CGI is not a language: it is a protocol used to access database information. For example, a WBM application may need to display the current number of incomplete work orders. This data could be stored in the database of a proxy workstation. A CGI application written in Java or Perl can be used to query the database and format an HTML page to display the information.

Java is rapidly emerging as an important tool for managing an intranet through a WBM system. Java is an interpretive programming language. Java code is not compiled before run time but is interpreted by a Java Virtual Machine (JVM) at run time. JVMs are available for many processor environments and are included with some Web browsers, such as Netscape Navigator and Microsoft Internet Explorer, enabling these browsers to execute Java code.

While Java can be used to build standalone applications and does not have to be scripted for Web-accessible applications, the importance of Java for Web-based management systems is its built-in Web-enabled capabilities. Applets, self-contained Java programs, are sent from the server to a browser-equipped client where they are run locally (Figure 1). Other CGI scripting languages such as Perl, another popular language used for Web-based applications development, run on the server. The back and forth nature of Perl has the drawback of consuming processing resources and network bandwidth, which can inhibit the application's performance.

Figure 1: Java applets work in one direction: server to client.

Applets have browser-imposed security restraints that limit access to network resources and prevent access to local system resources, such as memory and disk. Thus, applets can be run on the receiving machine with minimal risk while ensuring network security.

Java applets, which can be served from both proxy and embedded Web-server WBM implementations, can be used for:

§  Displaying dynamic graphs that interpret network operations

§  Illustrating complex situations, such as interactive views of chassis hubs or modules of a stackable hub

§  Displaying real-time data that are updated from polling and traps

§  Adding graphics, including animation

Because Java can produce applications that are portable across Unix, Windows, and other environments, a JVM can be embedded in an end device and the device agent can then execute Java code. This capability enables code to be ported across other devices with JVMs. A key facet of this capability is that the code can be distributed dynamically, ported from a management proxy to devices, and ported between devices or components within a device (Figure 2).

Figure 2: An embedded agent with a Java Virtual Machine.

The use of Java within an embedded agent can increase management capabilities by enforcing policy-based management or security rules. For example, assume that an administrator wants to enable remote network access between 6:00 a.m. and 6:00 p.m. Traditional methods, such as SNMP sets, have been used to enforce such restrictions but at the cost of increasing network traffic and with the difficulty of scaling to handle thousands of devices. While a device-resident agent could perform this function, agent releases are typically infrequent. An embedded agent with a JVM could independently create the scheduling routine and dynamically distribute the policy to restricted devices, bypassing the need for an agent release.