Introduction

An integral part of IBM®'s Business Process Management (BPM) portfolio, Monitor (hereafter called Monitor) is a comprehensive business activity monitoring (BAM) solution that provides a near real-time view of your business performance. BAM can provide visibility into the performance of business activities by processing events, calculating business metrics and presenting key performance indicators (KPIs) through business dashboards. BAM can also help when something goes wrong. When expectations aren't met, alerts can be delivered to make the organization aware of potential problems much earlier, allowing a directed action to be planned and carried out.

This article describes the highlights of Monitor V6.2. It covers everything from the dramatic changes to the out-of-the-box experience and added power at the fingertips of business users, the combination of KPI history, prediction and alerting that significantly improves decisions and actions, the accelerated and iterative process monitoring design for business analysts, the enhancements in ways of consuming events for solution architects, to the improvements to performance and scalability for IT. Finally, we'll point you to additional resources for more information.

Empower business leaders through Business Space powered by WebSphere

We begin our V6.2 tour with the end in mind. Monitor is usually deployed for the purpose of giving business users visibility and insight into the performance of their business, so that they can take informed action. The principal user interface to enable this visibility is the Monitor dashboard. Monitor V6.2 offers two dashboard deployment choices: Business Space powered by WebSphere and WebSphere Portal.

Monitor provides a dashboard known as Business Space, which provides users with the ability to manage their business through multiple interactive visualizations, known as widgets. Business users can dynamically build solutions by creating one or more pages and grouping them together into logical units known as spaces. A page is assembled by adding one or more widgets together in what is frequently referred to as a mashup.

Multiple WebSphere BPM products contribute widgets to the Business Space and as more products are installed, more of these widgets become available and in some cases can interact with each other. Monitor provides eleven distinct widgets, including a new one in V6.2 for showing history and predictions. Seven other widgets that come with all Business Space products complement the Monitor widgets by covering common tasks, such as viewing Web sites, presentations, documents, RSS feeds, gadgets and more.

As an alternative to using Business Space, Monitor also provides support for WebSphere Portal (Portal). Beginning in V6.2, Portal is no longer included with the product and must be purchased separately. The supported version of Portal is V6.1.0.1. Note that all of the Monitor-specific Business Space widgets are provided as portlets for use with Portal, with minimal differences between the two platforms.

Get started with the Better Lender Showcase

Also new in V6.2 is the Better Lender Showcase, shown in Figure 1. The showcase provides an out-of-the-box demonstration of the Business Space capabilities provided by Monitor. After installing the product, users can add the Showcase with a single click from the First Steps dialog. This configures all the necessary pieces, sends event data, simulates historical data, and creates a business space that has multiple pages and widgets ready for use. Users can then explore this space to familiarize themselves with its capabilities in a hands-on fashion. Included in the space is Getting Started documentation that walks the user through a scenario to help familiarize them with the business space. The new showcase means that, with just a few clicks, users can begin to experience Monitor V6.2. Figure 1 shows the first page of the Better Lender Showcase.

Enhancements to Monitor widgets

Monitor V6.2 has added new capabilities and changes to many of the existing widgets. A new user interface has been introduced which standardizes the look and feel of user interface controls and is being adopted across all products integrated in the Business Space. In addition, the Business Space has been significantly enhanced since V6.1 to add usability improvements like the grouping of pages into spaces, templates that provide predefined spaces for solutions, quick page and space switching, tab reordering, widget height and width customization, and enhanced page design through improved drag and drop support.

As shown in Figure 2, Monitor V6.2 has added a frequently requested feature for viewing months by their name instead of number (for example, October instead of 10) to the Reports and Dimensions widgets for improved dimensional analysis.

V6.1.2 also introduced significant enhancements to the Instances widget, which can now be customized across multiple models and multiple monitoring context levels. This allows for customized display of columns in drill-down and cooperative mode scenarios, including ordering or hiding of drill-down columns. The widget was further enhanced with the ability to export spreadsheet data in Comma Separated Value (CSV) format.

KPIs were also improved to add custom colors, and a custom palette for range color definitions was introduced. A feature known as range templates was added for quick definition of KPI ranges. These templates can be used across many KPIs. V6.2 also adds the ability to specify custom range colors in the monitor model editor in the Monitor Toolkit.

Monitor widgets provide monitoring of aggregate and instance specific data, and enhancements have been added to allow for drill-down scenarios that go from aggregate to instance data to help with issue resolution. Figure 3 illustrates how it you can now drill to Instances from Reports, Dimensions and KPI widgets, for example to show all the orders from new customer that contribute to a customer growth KPI.

Another cross-widget capability added in V6.2 enables cooperation between the Diagram widget and the Human Tasks widget using dynamic filtering. This allows users to click on points of agility within an aggregate diagram (typically human steps within a process diagram), and causes the Human Tasks widget to show only instances of the clicked human task. An example usage would allow a manager to click on the approval step in a loan approval process diagram and the Human Tasks widget would then show all of the approval tasks. A user can then perform actions on the task, such as claiming it to work on.

You can also click on a row in the Instances widget, which causes the Diagram widget to update to show an instance-specific diagram, and then click within that diagram to see the specific human task instance. In an example scenario, a user clicks on an individual loan application, and the diagram widget shows the processing state of that particular loan application. The user can then click on the Underwriting step in the diagram and the Human Tasks widget will show the specific Underwriting for the selected loan application. Figure 4 illustrates the two-step interaction between all three widgets.

Enhance agile decision making

Three major new capabilities have been added to Monitor V6.2 to support agile decision Making: KPI History, KPI Prediction and Dynamic Alerts. KPI History automatically captures the value of KPIs over time, and is complemented in the dashboard by a new KPI History and Prediction widget that analyzes KPI values over time. KPI Prediction builds on top of KPI History by allowing business users to define prediction models, then visualize these predictions in the same new widget, seeing the prediction trends in context of the history. Using the KPI History and Prediction widget, users can also specify the time period through simple controls that allow for powerful analysis over time periods.

Finally, Dynamic Alerts empower the business user to specify when, whom, and how to alert when situations occur. These alerts can be defined dynamically by users in the dashboards without any IT involvement. This final feature builds upon the prediction capability, because alerts can be defined to trigger based on predicted values. This combination of history, prediction and alerting enables significantly improved decision making to enable rapid and informed responses to business situations.

KPI History

When a KPI definition is deployed, or when a user dynamically creates a KPI through the dashboard, Monitor V6.2 begins capturing the historical values of the KPI on an hourly basis. It is even possible for the business user to retroactively calculate history for a KPI in cases where the existing data can be re-analyzed over a period of time. Furthermore, the history can be managed by an administrator who is able to import or export history.

The new KPI History and Prediction widget provides a time-based chart for analysis of the history data. Like the other widgets, it can be added to any space or page and configured using a simple business user interface to choose which KPI to display. Users can specify default time ranges and data frequency, for example one year of daily data points. Once the chart is configured to show this default, users can interact with it in the viewing mode to change the time range and data frequency on an ad-hoc basis, which allows for quick analysis of different time periods. The widget includes convenient controls for frequently used time windows such as one week, one month, month to date, year to date and more. Users can also specify any start and end date for arbitrary time ranges. They can display a tabular view of the data alongside the chart, and clicking data points in either the table or chart highlights them in both, allowing for easy cross-identification of data. Other usability features include hover information to provide data values, range names and values and targets. Figure 5 shows an example of the combined KPI History and KPI Prediction in a widget.

KPI Prediction

Monitor V6.2 also provides the ability to make predictions based upon this historical data. Predictions of KPI behavior enable improved decision making and faster reaction to problems or opportunities. Business users can define prediction models simply in the dashboard, specifying how far into the future to predict, and how frequent the prediction points should be. Additionally, prediction models can take into account cyclical patterns, such as a seven day weekly cycle or a quarterly cumulative sales trend. KPI predictions are made based on the KPI's past performance, current trends, acceleration and deceleration, and the optional cyclical pattern. For example daily predictions can anticipate low sales on weekends or peaks at quarter end.

Users can toggle the display of the prediction patterns on or off using the same widget that shows the history, whereby predicted data values are added to the chart and the table. When configuring the widget, users can select which of the predictions to show (multiple predictions can be defined for a KPI). Figure 5 shows an example of the KP History and Prediction widget, displaying history and prediction for a daily sales KPI, with the weekend pattern shown clearly in both the history and the prediction.

Dynamic Alerts

In previous releases, Monitor developers have been able to define business-relevant situations in the monitor model, and users have been able to subscribe to these modeled alerts. Although powerful, the need to involve an IT user and a deployment cycle for any changes to the alert definitions has restricted the scenarios in which this can be used. This capability has now been made dynamic in Monitor V6.2 with the introduction of Dynamic Alerts, which enables business users to define their own alert situations, alert content, and alert delivery mechanisms, as well as who should receive the alerts and how.

Dashboard users can now set up alerts on any of their KPIs by defining the conditions on which to be alerted. Multiple conditions can be combined, so users can be alerted, for example, when response time and call volume both increase above target levels at the same time. Users can receive a default notification or can customize the subject and body of the message, and may include other KPI details. Taking the call center example, this allows for receiving data about how many staff were working at the time of the call volume alert, as well as the call volume number, target and range expectations. When defining an alert, the user can also choose to share it with others who can receive notification, or if authorized can edit the alert definition itself. It is also possible to specify the frequency with which to be alerted; for example whether to receive a single notification or repeated ones.

Predictions can also be used in alert conditions, enabling users to create notifications that will be sent when predictions indicate problems or opportunities. This combination of historical analysis, future prediction, and dynamic alerting enables significantly improved agile decision making. As an example, consider a cumulative quarterly sales KPI where the historical trend shows that the final quarter needs to end strongly. The prediction model can be configured to show the quarter end projection, and alerts can be put in place if that projection is above or below target. The alerts can include information about other metrics that affect the sales figures, and users can make changes based upon the data presented. Figure 6 shows an example of a dynamic alert definition.

In conjunction with the new Dynamic Alert capability, Monitor includes a new Alert Manager widget which supersedes the Alerts Subscriptions widget in previous releases. With Alerts Manager, users can see their personal and shared alert definitions, with filtering to simplify the list and controls for editing, copying or deleting alert definitions.

Expand event sources

We now transition away from highlighting the features directly visible to business users in dashboards, and begin to focus on capability that can be used by other user roles to implement business requirements.

The value you get from monitoring increases as you gain broader and richer insight into business critical activities. V6.2 provides enhancements for monitoring the applications and activities running across your business via events from sources including WebSphere Process Server, WebSphere Business Services Fabric, FileNet P8, WebSphere MQ, WebSphere MQ Workflow, WebSphere ESB and WebSphere DataPower SOA Appliances. The following sections highlight specific Monitor V6.2 enhancements for better integration with WebSphere Process Server, WebSphere Message Broker and WebSphere Business Events, and describe new ways to publish events to be monitored.

Improved monitoring for previous versions of WebSphere Process Server

WebSphere Process Server (hereafter called Process Server) continues to be the most popular event source for Monitor customers. The Monitor Toolkit is well integrated with WebSphere Integration Developer (hereafter called Integration Developer) to provide a consistent development experience for processes and for the monitor models associated with those processes. If your project involves monitoring previous versions of Process Server, you can make use of a new capability that allows you to generate a monitor model in V6.2 that can monitor applications created in Integration Developer V6.0.2 or V6.1. This feature is packaged in the form of new Monitor Toolkit plug-ins that can be added to these previous versions of Integration Developer. You can also iteratively develop applications with the earlier version of Integration Developer while developing the monitor models in the new version of the Monitor toolkit. You can keep your monitor model synchronized with your application while both are being iteratively developed.

Improved control over events from WebSphere Message Broker

In conjunction with the timing of the release of Monitor V6.2, WebSphere Message Broker (hereafter called Message Broker) has been enhanced (as part of V6.1.0.3) to improve integration between the two products. These changes are also compatible with Monitor V6.1.2. Several improvements to the monitoring of message flows are introduced:

• You can now use the Message Flow Editor in the Message Broker toolkit to configure monitoring. In addition to the administrative commands provided in earlier versions, authors of message flows now have more control over which events to emit, and what content should be provided in the body of events. When the message flow is deployed from the workbench, it automatically emits events from the active event sources. Note that use of administrative commands may be preferred for scenarios where redeployment of message flows is not possible. Figure 7 shows an example configuration of monitoring properties as part of the Message Flow Editor.
  
  Figure 7. Message Flow Editor - Monitoring Configuration
  
  
  
• A new event type, terminal event, enables visibility to any significant occurrence inside a message flow. A terminal event can be emitted from any terminal on any node. Events can be emitted each time a message passes through the terminal. Note that the three input node event types (entry, exit, and failure) introduced in V6.1.0.2 are also supported.
• You can now capture complex data from the message payload and place it into the business events sent to Monitor. XPath is used to express the data to be captured. As shown in Figure 8, you can optionally author these expressions using the XPath expression builder provided with the Message Broker toolkit.
  
  Figure 8. Using the Message Broker XPath expression editor to specify event content
  
  
  Once you've defined the configuration for event emission, you can export XSD schema files for a message set from the Message Broker toolkit. These schema files can then be imported to the Monitor toolkit for use in creating monitor models that consume the Message Broker events.
• Message Broker events now have enhanced correlation properties. The values for local, parent, and global transaction correlators are saved in the Environment tree, and are used by all events. This enables scenarios like relating a message flow with a related business process executing Process Server.
• Finally, the mqsireportflowmonitoring and mqsichangeflowmonitoring administrative commands have been enhanced to operate on individual event sources from the command line.

Refer to the Message Broker Information Center, for more information on emitting events.

Bi-directional event flow with WebSphere Business Events

Beginning in Monitor V6.1.2, WebSphere Business Events (hereafter called Business Events) and Monitor provide an integration point whereby Business Events can send events to Monitor. This means that any event that can be sent to Business Events can also be forwarded to Monitor. In addition, complex events detected by Business Events can be sent to Monitor for correlation with other events and for providing visibility in dashboards. This integration is described as part of the developerWorks series "Business event processing with WebSphere Business Events"

Now, in conjunction with Monitor V6.2 and Business Events V6.2, events can also be sent from Monitor to Business Events. This enables advanced problem detection by feeding alerts to Business Events to detect hidden patterns in the alerts and indicate a potential business problem that might otherwise go undetected. For example, by feeding alerts from Monitor that predict client demand will exceed inventory, and correlating these to events from supply chain events, business users can gain information about potential stock shortages that was not previously available to them.

As shown in Figure 9, the Business Events authoring tool can be used to create filter rules that use the event structure that is emitted by Monitor.

Publish events via REST API

Monitor V6.1 added REST APIs for accessing the Monitor metadata and data. Monitor V6.2 now adds a REST API to publish events for consumption by Monitor. This new REST service, identified by the /rest/bpm/events URI, hides implementation details related to event infrastructure. The interface provides two event publish options as set by these content types:

• Use the text/xml content type to send a single XML event.
• Use the application/atom+xml content type to group a batch of events with multiple XML event payloads.

Publish events with the Web Services-Notification API

Monitor V6.1 enabled the use of the WS-Notification standard API to send events to Monitor by providing new support for XML events and event sequencing technology. A two-part developerWorks article Publishing event messages to IBM WebSphere Business Monitor V6.1 with Web Services Notification provides an explanation and examples for use with Java™ and with .NET™ environments. Monitor V6.2 adds to this, providing example configuration scripts and documentation to guide administrators in implementing the WS-Notification option with Monitor server. A Web service can publish a business event as a notification message, and the application server can stream the message to the input JMS queue for Common Event Infrastructure (CEI) and then on to Monitor.

Accelerate time-to-value across lifecycle

The following sections highlight new capabilities to move quickly from requirements for monitoring to realization in dashboards and actions. Improvements have been made that are relevant to business analysts, developers and administrators.

Interactively design, emulate and test processes and monitor models

Monitor V6.2 has been enhanced to enable a new integration point with WebSphere Business Modeler (hereafter called Modeler). For a select number of monitoring design and test scenarios, a business analyst can define processes and business measures in Modeler and deploy to Process Server and Monitor Server for emulation and testing with minimal IT intervention. This dramatically reduces the time between forming a requirement and validating the process and monitoring design in Business Space.

This new accelerated design is focused on scenarios in which human-centric processes are executing in Process Server, and the measures required in the dashboard are calculated from Process Server tasks and processes. Business analysts can easily add process and task measures using Modeler's business measure templates. Note that the use of Integration Developer is still required for implementation of services, and for more advanced process and monitoring requirements. You must also use Integration Developer to create project artifacts prior to the actual production deployment.

As shown on Figure 10, business analysts can select a process and choose the Modeler Test on Server action. This action then initiates a new Monitor service that consumes a Modeler model and creates the deployable artifacts, including J2EE projects and database schemas. These artifacts are then deployed to a Process Server and Monitor server that has been installed and configured by IT.

In addition, a business space with Monitor widgets is automatically created and deployed so that users can test the modeled processes and see business metrics, KPIs and dimensional views for the running processes without the extra steps of configuring a business space. Figure 11 shows a business space page including process interaction, form creation, task management and monitoring that has been automatically created from Modeler and deployed to a Process Server and Monitor server.

Reusable KPI libraries

To speed the selection of the KPIs most applicable to your business, Monitor V6.2 adds a new KPI wizard in the Monitor Toolkit. This wizard, shown in Figure 12, provides access to a library of over 800 open standard KPIs based on the APQC Process Classification Framework (PCF). The KPI library enables business analysts to select KPIs for various processes across several functions including financial management, human capital management, customer relationship management, and supply chain management.

Graphical debugger with event infrastructure integration

The Monitor model debugger was added in Monitor V6.1.2 as way to understand the operation of a monitor model and to identify and resolve model logic problems. You can send events through the integrated test client and step through the monitor model execution. Monitor V6.2 now provides full support for debugging events from any source, not only those events sent from the integrated test client.

Enable BAM on mobile devices and on the desktop

In Monitor V6.2, there are more ways than ever to access BAM information beyond the traditional dashboard environment. You can view mobile dashboards from your smartphone devices, see and interact with live data in your Microsoft® Excel® spreadsheets, and access BAM data in your instant messaging (Lotus® Sametime) and e-mail (Lotus Notes) environments.

Mobile dashboards

V6.1.2 introduced support for viewing and interacting with BAM data on your RIM Blackberry® device. In V6.2, support has been added for Apple®'s iPhone™ and iPod™ Touch devices. By simply pointing the device's Safari® browser at a Web application provided with the product, you can see all of your alerts, human tasks, and the values of your KPIs. This Web application has the look and feel of a native iPhone application, but without the requirement to install any applications locally on your device. Thus, even from the road, you can stay connected with what is happening in your business, including taking action, such as forwarding an alert, or reassigning a human task that is hosted in Process Server.

Microsoft Excel add-in

Earlier versions of Monitor included the ability to export data as an Excel spreadsheet. V6.2 adds the option to install a new add-in "ribbon" into Microsoft Excel 2007. This ribbon can be used to import Monitor data to a worksheet and keep it current (via auto-refresh). Using this ribbon, you can choose which metrics you want to see, and then import instance data for a specified monitor model, resulting in a worksheet containing a row for each monitoring context instance, and a column for each metric. Once this data is imported, you can perform whatever manipulation you want on it, including calculations and graphing, and this data automatically refreshes as the underlying data in the Monitor database is updated. You can also choose to import and work with live data related to alerts and KPIs. Furthermore, since the connection info is stored in the Excel document, you can send this spreadsheet to a colleague, and when they open it up, they too will see up-to-date information, rather than just a stale snapshot. Figure 14 shows BAM data in Excel.

BAM plug-in for Lotus Sametime

Another environment in which you might want to have access to BAM data is your instant messaging client. Monitor V6.2 introduces a plug-in for Lotus Sametime that lets you see your KPIs, alerts, and instance data, and act upon that data, such as forwarding an alert. This plug-in is integrated with the standard notification features of Sametime, so you can be informed immediately of an alert, rather than having to poll for it periodically. Starting with Lotus Notes V8.0, Sametime support is integrated, so you can use the Monitor plug-in there as well, in the Notes sidebar. Because users have access to BAM data directly within their e-mail and instant messaging applications, they can more easily react to problems in their business as part of their normal routine, and interact with others to address them, rather than having to open a separate Web browser periodically to check on the status of their business. Figure 15 shows BAM data in Lotus Sametime.

Improved tools for administrators

Monitor V6.2 now boasts several new tools to ease the job of IT administrators who need to set up and maintain a production environment. These include a new wizard in the administration console that helps set up a network deployment (ND) topology, support for recording events that have arrived for a given monitor model (and the ability to play back that recording if desired), and several new scheduled services that can be configured to automatically manage things like data archiving and pruning and cache refreshing.

Monitor topology wizard

The Monitor topology wizard greatly simplifies the process of configuring Monitor in a production environment. While a simple standalone profile is fine for testing and demoing monitor models, it's strongly recommended that any true production use of Monitor be deployed as part of an ND cell that is configured for high availability, to avoid any single point-of-failure. V6.2 adds a new wizard into the administrative console (Servers => WebSphere Business Monitor configuration) that helps with configuring the applications and resources required by Monitor. From a single dialog, you can see whether your cell is properly configured and follow links to address any items indicating a problem. You can use this wizard regardless of whether you choose to augment the same cell as your event source (such as Process Server) for Monitor, or whether you choose to place Monitor into a separate cell. Either way, the wizard helps you configure everything properly for workload distribution and high availability, dramatically reducing the complexity of setting up a production deployment of Monitor. Figure 16 shows the topology wizard.

Event recording and playback

In Monitor V6.2, administrators can record an event stream for later export or playback. This ability can be useful for problem determination, for example, if you want to debug why the model behaved a certain way, you can look at the actual events that led to the problem as input for the debugger. This can also make it easier to service the environment without having to provide live access to it to outsiders, such as consultants or other support staff. And, perhaps most importantly, this function can be critical to recovering from a system failure, making it possible to reprocess any events, such as those that might have arrived since the last good snapshot of the environment, that would otherwise have been lost forever. This record and playback capability is available in both the toolkit (in the context menu for the UTE server) and in a production environment (in the administrative console). Figure 17 shows the Recorded Events Management dialog.

Scheduled data services

Users of prior versions of the product may recall the optional scheduled Data Movement Service (DMS). In V6.2, several new scheduled data-related services have been added and consolidated into a single, easy to use and administer dialog in the administrative console. These scheduled services cover topics that have been discussed earlier in this article, such as when history values should be recorded for KPIs, when predictions should be made for KPIs, when dashboard-defined alerts should be evaluated, and when caches should be refreshed. One additional service not discussed so far is the ability to archive and prune old instances that may no longer be required. For example, say you only want to store data for instances active within the past year; this service will automatically archive such data to a file that can be imported to a warehouse if desired, and then purge it from the Monitor database, ensuring that you're only seeing data for the desired timeframe and keeping your history of data from growing too large.

These scheduled services are also synchronized and ordered, so that one doesn't run until another that it depends upon has run. This ability to easily see and administer all of the scheduled services for each monitor model is a powerful new capability for Monitor administrators. Figure 18 shows the Monitor Scheduled Services dialog.

Improve performance, scalability, and latency

Several key improvements were made regarding performance and scalability in V6.2. Significant optimizations were made to event routing, event processing, dashboard latency, and dashboard response time with a deep history of instance data.

Event throughput improvements

In the area of event routing, dramatic scalability improvements were made in the case where there are many event subscriptions (that is, many installed monitor models). In prior releases, each event that arrived at the CEI server was parsed once per subscription; in V6.2, CEI parses the Common Base Event (CBE) XML only once, regardless of how many subscriptions exist. This means that, with many subscriptions, and especially with large events, CEI can now route significantly more events per second to their appropriate monitor models before reaching CPU saturation. In addition, if even greater throughput is required than a single multi-CPU machine can deliver, CEI can be horizontally clustered across machines, so that the event routing workload is managed across the members of the cluster.

Another key improvement in Monitor V6.2 is the option to bypass use of JMS queues for getting events from CEI over to the appropriate monitor model, and instead use the Monitor database as the place to which the event is delivered. This significantly speeds up event processing for a monitor model, since the cost of assured persistent messaging is avoided (without any risk of losing the events), and the transactions are now single-phase commit, which run faster and with less overhead. Initial measurements have shown that a monitor model installed to a single machine on V6.2 with queue bypass can run as fast as what previously required two machines on V6.1.2. This improvement, coupled with the improved routing mentioned earlier, also reduces the cost of enabling event emission, such as from a BPEL application, especially when it is emitting its events synchronously. This queue bypass improvement can also improve reliability, since it removes messaging engines or cross-cell bus links as points of possible failure along the path of event delivery and event processing by monitor models. Now only the Monitor database needs to be available in order to deliver and process events. Note that this queue bypass technique is now the default option, but queue-based delivery will continue to be supported for backward compatibility.

Dashboard latency and load-time improvements

Dashboard users will now experience less latency in the data they see; that is, there won't be as long a delay between an occurrence and when the dashboard reflects that occurrence. With the new scheduled services mentioned earlier, you can control both KPI and cube caching in an efficient manner. These coordinated services can help reduce latency experienced in the dashboards, avoiding issues such as DMS occurring at the top of the hour, but cache refreshes occurring at the bottom of the hour (and thus a latency of 1.5 hours, rather than the intended one hour). Also, KPIs where history is enabled are computed automatically on the hour, so the cache will be refreshed as part of that, rather than waiting until the next dashboard request to update the cache, meaning users don't have to wait through the potentially expensive KPI computations.

Significant improvements were made to the speed of dashboard page loading when there is a large amount of data collected for a given model (where the number of instances is in the 10 million or greater range). V6.2 enables finer-grained control over the cache intervals for each KPI: these can now be set via the business space per KPI. This is particularly important for KPIs that have a wide filter, that is, KPIs that require computations across a large percentage of the instances collected for that model. And to accelerate dimensional analysis, V6.2 now supports optional Materialized Query Tables (MQTs), which effectively precompute and store the results of expensive queries (for cube measures), thus avoiding potentially long delays during cube navigation, such as a drill-down operation. These MQTs are refreshed as a coordinated scheduled service as well; and when the MQT refresh is happening, users will continue to see the cached data, instead of waiting for the potentially expensive refresh to complete.

Support for more platforms

Monitor V6.2 also broadens the list of supported operating systems and databases, bringing it more into synch with the platforms supported by the rest of the IBM BPM suite. Support has been added for the 32-bit versions of Windows® 2008 and Solaris® 10, and for 64-bit versions of the following operating systems: Windows 2003, Windows 2008, AIX® 6.1, HP-UX® 11iv3 for Intel® Itanium®, Red Hat® Enterprise® Linux (RHEL) 4.0 for Intel, RHEL 5.0 for Intel, SUSE Linux Enterprise Server (SLES) 9 for Intel and SLES 10 for Intel.

V6.2 adds support for Oracle® 11g, as well as for newer fixpack levels of previously supported databases, such as DB2® 9.1.5 and 9.5.2a. Support has also been added for Highly Available (HA) databases, including DB2 HADR and Oracle RAC. Now your database doesn't have to be a potential single point-of-failure, but rather it can fail over to a mirror and keep running with no downtime (in-flight transactions are automatically retried after the failover, with no data loss).

Summary

WebSphere Business Monitor V6.2 is a major release that offers significant new capability for monitoring your business processes and activities. This article gave you a high-level tour of many of the new features and enhancements. The following table summarizes the new highlights in Monitor V6.2 across eight primary categories.

Acknowledgements

The authors would like to thank the entire Monitor V6.2 team for their innovation and commitment, without them this work would not be possible. Specific acknowledgements to the team leaders: Christina Watkins, Clayton Sims, Curtis Miles, Dan Willey, David Enyeart, Jim Thorpe, Ke Jia Li, Latha Sivakumar, Nick Metianu, Paritosh Patel, Thomas Burke, Varadarajan Ramamoorthy and Wilfred C Jamison.

Resources

• WebSphere Business Monitor education: A course list of WebSphere Business Monitor classes
  
  
• Browse the WebSphere Business Monitor V6.2 Information Center: Get complete product documentation.
  
  
• The possibilities of BAM everywhere (developerWorks 2007): This IBM Business Process Management Journal column explores a range of ways to benefit from information collected by Monitor.
  
  
• What's New in WebSphere Business Modeler V6.2 (developerWorks 2007): Learn more about the new design-to-deploy features for processes and process monitoring.
  
  
• Refer to What's New in WebSphere Process Server V6.2 (developerWorks 2007): Read about WebSphere Process Server's use of Business Space and many other exciting features in V6.2.
  
  
• WebSphere Message Broker 6.1 Information Center: Get more information about controlling events that can be emitted for consumption by Monitor.
  
  
• developerWorks articles related to Monitor: Check out other developerWorks articles on WebSphere Business Monitor for a wealth of how-to information and samples.
  
  
• developerWorks BPM zone: Get the latest technical resources on IBM BPM solutions, including downloads, demos, articles, tutorials, events, webcasts, and more.
  
  

About the authors

Eric Wayne is Lead Architect for Business Activity Monitoring in Business Process Management. He leads the Monitor Development team and is a core member of the IBM Software Group Architecture Board. You can reach Eric at ewayne@us.ibm.com.

John Alcorn is the lead architect for the IBM Business Activity Monitoring (BAM) platform. He has worked as a software engineer with IBM for 15 years, with more than ten years on WebSphere products, including roles in both product development and software services. John has been a technical leader with the WebSphere Business Monitor product for 4 years, and works closely with the wider IBM WebSphere Business Process Management (BPM) team.

John is IBM-certified in XML technologies, SOA technologies, and in multiple WebSphere products, and is Sun™ certified in Java™ programming. He currently manages a team of developers at the Research Triangle Park lab in North Carolina. You can reach John at jalcorn@us.ibm.com.

Richard Johnson is Lead Architect for the WebSphere Business Monitor dashboards and is a Certified Consulting IT Specialist. Over the past ten years in IBM he has worked in various areas within Software Group, including test, development, beta program management, and IBM Software Services for WebSphere, spending much of the time in customer facing positions. These roles have spanned multiple products, technologies, and geographies - from the mainframe and CICS to distributed WebSphere platform products. He worked for IBM UK before moving to IBM US in 2006.

You can reach Richard at rj1@us.ibm.com.

Comments (Undergoing maintenance)

Back to top

Trademarks  |  My developerWorks terms and conditions

Table of contents

• Introduction
• Empower business leaders through Business Space powered by WebSphere
• Enhance agile decision making
• Expand event sources
• Accelerate time-to-value across lifecycle
• Enable BAM on mobile devices and on the desktop
• Improved tools for administrators
• Improve performance, scalability, and latency
• Support for more platforms
• Summary
• Acknowledgements
• Resources
• About the authors
• Comments

Local resources

• IBM Innovation Center - Austin, TX
• IBM Innovation Center - Chicago, IL
• IBM Innovation Center - Dallas, TX
• IBM Innovation Center - San Mateo, CA
• IBM Innovation Center - Waltham, MA
• Technical Briefings in North America

My developerWorks community

Dig deeper into WebSphere on developerWorks

Rate a product. Write a review.

Write a review for Rational products today!

Special offers