BACKGROUND OF THE INVENTION
The present invention relates in general to enterprise information systems. More particularly, the present invention relates to methods and apparatus for optimizing the delivery of data to a device.
Timely intelligence is critical to proper business decision making. Collecting and quickly analyzing information, however, is often difficult. In a dynamic marketplace, delayed delivery of intelligence can reduce both its reliability and relevancy. Substantial delays can even make the information, often acquired at considerable cost, completely worthless.
Intelligence is often extracted from OLTP applications and from specialized databases, called operational data stores, OLTP (online transaction processing) applications are those enterprise systems that manage a company's basic transactions, such as order entry and billing systems. Since they are commonly optimized for reading and writing, and not for querying, extracting information from an OLTP application can be sufficiently complex to require additional user training. Furthermore, while operational data stores generally archive OLTP information in a format for querying, they also generally do not maintain important historical information. For instance, an operational data store may store a current checking account balance, but not the individual daily balances over the previous month.
In addition, the queries themselves often take substantial time to execute, producing only static snapshots of the information. Observing the dynamic aspect of changing information is therefore difficult. A possible solution may be to sequentially execute and aggregate a series of queries. However, this solution can be both inefficient and ineffective, since manual manipulation still often delays the information delivery. That is, these queries must be first manually aggregated and summarized, before intelligence reports can be generated and delivered.
ETL (extraction, transformation, and loading) systems help by extracting, transforming, and aggregating the information. But latency inherently found in most distributed networks, coupled with considerable manual intervention that ETL systems often require, mean that critical information can still be received late.
Automatic notification is a possible solution. Many analytical products such as business intelligence (BI) and online analytical processing (OLAP) systems are capable of monitoring, scheduling, and broadcasting alerts via email or pager. These systems, however, cannot generally assure that the intended recipient promptly receives and reads the message. This presents a significant problem for time-sensitive information, where minutes or even seconds can make a difference.
To facilitate discussion, FIG. 1 shows a simplified functional diagram of distributed information architecture. The diagram can be divided into an enterprise data layer 160 and a client layer 162. Enterprise data layer 160 comprises elements that are primarily focused on accumulating, processing, and transforming operational data. Client layer 162 comprises elements that are primarily focused on rendering the processed data for a user.
OLTP (online transaction processing) applications 152 are commonly coupled to each other, as well as to other enterprise applications, through a dedicated messaging and queuing application (MQ), such as IBM's MQSeries. MQ provides an efficient communication channel for these applications, by storing and forwarding data messages, in a manner that is similar to email.
Commonly coupled to each OLTP application 152 is operational data store 154, such as an Oracle database. Through an API (application programming interface), transactional data can be transferred between the OLTP application and the database. Operational data store 154 consolidates that data from multiple sources and provides a near real-time, integrated view of volatile, current data. Since its purpose is to provide integrated data for operational purposes, operational data store 154 primarily has add, change, and delete functionality.
In order to conduct meaningful analysis, this information is often further placed in a more stable environment, optimized for random querying. ETL system 155 extracts the information from the appropriate data store 154, transforms and combines the data based on pre-defined constraints, and subsequently loads the data into data warehouse 156. A popular ETL technique, developed by Sagent, is the use of data flows.
Data flows are a series of rule-enabled transformations that are connected in data pipelines. They handle the tasks of joining, merging, comparing and splitting data and permit the separation of data into different logic paths, each of which can be further combined and split off to create more complex transformation sequences.
ETL data extractions often occur by either a bulk or a trickle method. In the bulk method, periodic snap shots of data in operational data store 154 are extracted and uploaded into data warehouse 156. This commonly occurs as a large batch file scheduled during a low system utilization period. In the trickle method, changes in operational data store 154 are continuously uploaded, or “trickled” into data warehouse 156. These updates are therefore frequent, smaller, and more current than in the bulk method. As in the case of OLTP 152 systems, ETL 155 can also use the MQ for data extraction.
Once the data is in data warehouse 156, it is available for OLAP 158 (online analytical processing). OLAP enables trained users to perform ad hoc analysis of data in multiple dimensions, such as with an OLAP cube. OLAP cubes provide multi-dimensional views of data, querying, and analytical capabilities. Furthermore, many OLAP products can schedule, run, publish, and broadcast reports, alerts and responses over the network, email, or personal digital assistant. Users often access OLAP 158 by thin client 162. Thin clients are applications that generally are integrated into the underlying client device, and generally require minimal modification. For instance, a thin client can be browser with a Macromedia Flash module installed.
Although OLAP analysis can provide valuable insight about business operations, critical information is often received late, even with automated reporting. Automated OLAP reporting often only has access to the information within data warehouse 156, which can be several processing stages behind OLTP 152. This delay can be substantial, reducing the information's value. Furthermore, these reports are often only static snapshots of information in data warehouse 156.
For example, a NASDAQ broker places an order into an OLTP 152 application called an electronic communications network, or ECN. The ECN matches customer buy and sell orders directly through the computer. In this case, an order to buy 100 shares of ABC at $18.75 was entered. This open order is stored in the ECN operational data store 154, subsequently extracted by ETL 155, and analyzed by OLAP 158. If the buy order amount is the then highest in the ECN, OLAP 158 forwards the information to thin client 164, NASDAQ quote montage, where it is immediately displayed on the familiar stock market ticker tape. And although this system delivers stock information to individual brokers with reasonably small latency, it is also not easily modified. The NASDAQ application is custom designed for the specific purpose of enabling stock trading. As such, it would be difficult to display additional data on the stock ticker, such as non-financial information, without substantial additional programming.
In view of the foregoing, there is desired a method and apparatus for optimizing the delivery of data to a device, in which relevant information is received in a timely manner, and in which that data is rendered in a dynamic format.
SUMMARY OF THE INVENTION
The invention relates, in one embodiment, to method for optimally monitoring a set of data elements for an event. The method includes determining a set of data elements, each of which further comprises a state. An event is commonly a change in this state. The method further includes creating a rule for monitoring the event, wherein the rule comprises an event identification portion resident on a first computer and an action portion resident on a second computer. The method also includes monitoring the set of data elements for the event based, in part, on the event identification portion, and determining if the event has occurred. If so, the method further includes sending the set of data elements to the second computer. The method also includes executing a set of actions, based in part, on the action portion. Thereafter, the method further includes monitoring the set of data elements based, in part, on the rule.
In another embodiment, the invention relates to an apparatus for optimally monitoring a set of data elements for an event. The apparatus includes a means for determining a set of data elements, each of which further comprises a state. An event is commonly a change in this state. The apparatus further includes a means for creating a rule for monitoring the event, wherein the rule comprises an event identification portion resident on a first computer and an action portion resident on a second computer. The apparatus also includes a means for monitoring the set of data elements for the event based, in par, on the event identification portion, and a means for determining if the event has occurred. If so, the apparatus further includes a means for sending the set of data elements to the second computer. The apparatus also includes a means for executing a set of actions, based in part, on the action portion. Thereafter, the apparatus further includes a means for monitoring the set of data elements based, in part, on the rule.
These and other features of the present invention will be described in more detail below in the detailed description of the invention and in conjunction with the following figures.
BRIEF DESCRIPTION OF THE DRAWINGS
The present invention is illustrated by way of example, and not by way of limitation, in the figures of the accompanying drawings and in which like reference numerals refer to similar elements and in which:
FIG. 1 illustrates a simplified functional diagram of distributed information architecture;
FIG. 2A-B illustrate, in accordance with one embodiment of the present invention, simplified functional diagrams of the active data platform;
FIG. 3A illustrates, in accordance with one embodiment of the present invention, a simplified functional diagram displaying the active studio application;
FIG. 3B-E illustrate, in accordance with one embodiment of the present invention, several simplified de-normalized tables as would be used in an active data cache.
FIG. 4 illustrates, in accordance with one embodiment of the present invention, a simplified functional diagram displaying an event engine;
FIG. 5 illustrates, in accordance with one embodiment of the present invention, a simplified functional diagram displaying components of the rich client and active data cache.
FIG. 6 illustrates, in accordance with one embodiment of the present invention, a simplified functional diagram displaying components of the message center;
FIG. 7A-E illustrate, in accordance with one embodiment of the present invention, exemplary screen displays.
FIG. 8A-B illustrate, in accordance with one embodiment of the present invention, an active collaboration server advantageously employed to allow impromptu graphical collaboration between IM users;
FIG. 9A-B illustrate, in accordance with one embodiment of the present invention, simplified functional diagrams of a rules-based monitoring engine with alert capability;
FIG. 10A-C illustrate, in accordance with one embodiment of the present invention, simplified pseudo-code as would be used by IM client.
FIG. 11 illustrates, in accordance with one embodiment of the present invention, a simplified functional diagram of the active data platform with elements of the notification and escalation process;
FIG. 12 illustrates, in accordance with one embodiment of the present invention, a simplified diagram of a common instant messaging client displaying a notification message; and,
FIG. 13 illustrates, in accordance with one embodiment of the present invention, a simplified functional diagram of an active data platform in which distributed rule evaluation is implemented.
DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS
The present invention will now be described in detail with reference to a few preferred embodiments thereof as illustrated in the accompanying drawings. In the following description, numerous specific details are set forth in order to provide a thorough understanding of the present invention. It will be apparent, however, to one skilled in the art, that the present invention may be practiced without some or all of these specific details. In other instances, well known process steps and/or structures have not been described in detail in order to not unnecessarily obscure the present invention. The features and advantages of the present invention may be better understood with reference to the drawings and discussions that follow.
In accordance with one embodiment of the present invention, an active data architecture is advantageously employed to facilitate the presence-aware delivery of timely data. That is, determining the type of device available to the user, and delivering the information in a format within the device's capabilities. In accordance with another embodiment of the present invention, information can be continuously extracted from data sources without the need for manual queries, or additional user training. In accordance with another embodiment of the present invention, information may be dynamically delivered in order to render streaming real-time or near real-time data and/or analytical results, whether graphically, textually, audibly, or otherwise.
Referring now to FIG. 2A, a simplified functional diagram of an active data platform is illustrated. Active data platform 210 comprises a suite of applications that allow a user to actively monitor continuously changing information, throughout enterprise data layer 140, according to a pre-defined set of rules. A rule may be a pre-defined procedure comprising the events to be monitored, and the subsequent action or transformation that may be required. These rules enable active data platform 210 to extract, combine, and transform data from various systems, as well as the MQ messages transferred among the applications, in a substantially contemporaneous manner. This processed information can then be transmitted in substantially near real-time to devices in client layer 142. For example, processed information can arrive at a client layer 142 device within 2-10 seconds of an event or transaction, or even sooner. Depending on the type of client device, this processed information can be received as a continuous data stream, a static report, or an alert. The processed information can also be stored in a data warehouse and compared with future information.
As in FIG. 1, OLTP 102 generates transactional information. Operational data store 104 consolidates that data from multiple sources and provides a near real-time, integrated continuously updated view of data. ETL system 105 extracts the information from the appropriate data store 104, transforms and combines the data based on pre-defined constraints, and subsequently loads the data into data warehouse 106 that is optimized for querying, or random data retrieval. The ETL extraction can occur in either a bulk or a trickle method. Once the data is in data warehouse 106, it is available for OLAP 108, or online analytical processing applications. OLAP enables trained users to perform ad hoc analysis of data in multiple dimensions, such as with an OLAP cube.
Through the use of browser technology, such as DHTML, thin client 114 and a rich client 282 provide the user with enhanced visualization functionality, such as the capability to graphically render streaming data. Thin client 114, in one embodiment, is called active viewer, and comprises a browser. Rich client 282, in another embodiment, is called active viewer pro, and comprises a browser with additionally installed software. DHTML is a combination of HTML, style sheets, and scripts, and allows web-pages to be animated. In addition, rich client 282 further allows for enhanced functionality, such as pen-based user input, secure instant messaging, robust printing capability, and the use of skins. Skins are configuration files that allow modification of the browser appearance. Messaging client 115, such as a SMS cell phone, or an instant messaging program, such as Yahoo Messenger or Windows Messenger, provides alert capability. These alerts, as well as static reports, can also be received through thin client 114, such as a browser. Other 280 represents those enterprise applications that have been programmatically coupled to active data platform 210, as a client.
In one embodiment, client applications in client layer 142 can reside on a Tablet PC. The Tablet PC represents the evolution of the business notebook personal computer. It marks a new direction for notebook PCs, adding pen-based technologies to existing notebook functionality and enabling Windows XP-compatible applications to take advantage of various input modes, such as pen and voice-based data. The Tablet PC will be among the most versatile PCs ever made, with high-powered and efficient processors, lightweight designs, and, for man, built-in wireless capability. All Tablet PCs will have keyboards, some permanently attached and others detachable. But the Tablet PC is more than just a great notebook. The real value of the Tablet PC is that it makes it possible for information workers to bring the power of their business PC into more places than ever before, and they can do it with the same software they use on their office PC or notebook today.
For example, betting transactions would be normally stored in a casino's game transaction system. Wanting to be extra-attentive to high-rollers, a casino would configure the active data platform to observe for any customer whose betting exceeds a specified threshold, upon which an alert would be sent to the closest pit boss. The pit boss would locate these customers and personally oversee that exceptional customer service was provided.
FIG. 2B illustrates, in accordance with one embodiment of the present invention, a simplified functional diagram displaying elements of the active data platform 210, shown in FIG. 2A. Enterprise links 254 comprises adaptors that allow active data cache 256 to be coupled to the applications in enterprise data layer 140. Furthermore, these adaptors can also selectively filter and transform incoming data for active data cache 256. For instance, adaptors in enterprise link 254 can allow active data cache 256 to connect to enterprise applications (i.e., SAP, Siebel, and PeopleSoft), databases (i.e., Oracle, Informix, Sybase, DB2, and SQL Server), legacy applications (i.e., IBM S/390 mainframes), OLTP (i.e. order entry, billing, payment, etc), and MQ applications (i.e., IBM MQSeries, Vitria, Tibco, etc.).
Active data cache 256 is the core repository for the active data platform. Unlike traditional databases, such as operational data store 104 and data warehouse 106, active data cache 256 stores data in a de-normalized format. Normalization is the practice of optimizing the database structure into multiple tables, eliminating data redundancy and improving sealability and maintainability.
Since active data cache 256 is primarily focused on automatic reporting, and not on querying, it generates substantially de-normalized tables representing each report or data stream. By logically placing these de-normalized tables near client layer 142, in a non-obvious fashion, fewer processing stages are required and latency is substantially reduced.
Database tables are not commonly de-normalized, since data is duplicated. In general, duplicated data tends to increase the likelihood of data corruption during updates. That is, multiple separate entries no longer match each other. Furthermore, de-normalized tables tend to be more difficult to query than normalized tables, since each individual data record may be substantially larger, and hence take longer to search. This is even more so if the table is not properly indexed. However, since ad-hoc querying is not a substantial requirement, and table joins are not necessary, normalization enables database tables within active data cache 256 to be optimized for the efficient creation of pre-defined reports and alerts. In one embodiment, a single de-normalized table is the source for each report.
Active design process 290 is powerful, intuitive suite of applications that allow a user to create rules, map onto message queues, manage security, define data objects, and configure enterprise link 254 processes and sources, based on simple drag-and-drop operations for the data streams, reports, and alerts. In addition to active design process 290 in active data layer 241, there is a corresponding active design process client in client layer 142, such as a web browser. In one embodiment, active design process 290 comprises an application entitled architect. Unlike traditional real-time business intelligence applications, active design process 290 allows a user to dynamically add new data sources, or create new transformations and reports, without changing the system as a whole. That is, unlike the NASDAQ stock trading example previously shown, which is not easily modified, active design process 290 can easily tailor the active data platform to new applications and changing corporate infrastructures.
Active report engine 274 converts the raw information in active data cache 256 into charts, tables, and graphs for display on a device in the client layer 142. Active studio 291 is a presentation editor that allows a user to easily create and format specific graphical and textual active views of data, based on pre-defined templates. Common views are tables, pie graphs, bar graphs, line graphs, scatter graphs, columnar reports, form reports, geographic (map) views, crosstabs, and Excel integration. For instance, a user can create an hourly report in which gambling customers are aggregated into different groups based on the value of their accumulated betting that day. The user selects a bar graph for the report, and changes the fill color and font type to the user's preference. These views can be substantially active, changing in near real time as the underlying data changes, or can be static, like a printed report.
Event engine 257 monitors the information in active data cache 256 for pre-determined changes, or events. Once an event occurs, such as a rising temperature in a chemical processing plant, message center 272 is notified, subsequently forwarding the appropriate message to messaging client 115, thin client 114, or rich client 282.
In one embodiment, message center 272 tracks the presence of client devices, as well as the status of users themselves, in order to substantially assure message delivery. This is in contrast to traditional messaging systems that cannot determine if messages are, in fact, delivered to intended recipients. Presence determines if the device is available on the network. Status determines if the intended recipient is available to read and respond to the instant message. Common status categories include descriptive information, such as “out to lunch”, “busy”, “away from desk”, “important meeting”, etc. For instance, the device may be turned on and available, but the intended recipient is in an important meeting and does not wish to be disturbed. The intended recipient sets the status to “important meeting” in order to notify message center 272, as well as any other subscribed user, that he will probably not respond to the alert until after the meeting. Message center 272 would then choose an alternative available recipient, based on an established notification rule.
For example, a casino's active data platform attempts to notify the nearest pit boss of a particularly prolific gaming customer. However, since the pit boss has a “busy” status, the casino manager is alerted instead, and subsequently walks over to greet the customer.
Enterprise security system 258 represents a company's existing security infrastructure, such as Steel-Belted Radius by Funk Software. Since operational information may be sensitive, active data cache 256 can be integrated into enterprise security system 258. In this way users are authenticated before information is transmitted to client layer 142 devices.
FIG. 3A illustrates, in accordance with one embodiment of the present invention, a simplified functional diagram displaying the active studio 304 application of active data platform 210, as shown in FIG. 2B. In addition to active studio 304 application in active data layer 241, there is also a corresponding active studio 304 client in client layer 142, such as a web browser.
FIG. 3B-E illustrate, in accordance with one embodiment of the present invention, several simplified de-normalized tables as would be used in a casino's active data cache 256, as shown in FIG. 3A. Referring to FIG. 3B, a de-normalized table is shown describing payouts at specific gaming tables on the casino floor. For example, table 1, whose current pit boss is Jones, and whose current dealer is Brown, has a payout of $2000. Referring to FIG. 3C, a de-normalized table is shown describing the current location of players in a casino. For example, player Bill Smith, who is a high roller, is currently at slot machine #1. Referring to FIG. 3D, a de-normalized internal table is shown describing active server user profile information. For example, a system user with ID#1 prefers to receive notifications and alerts through email. Referring to FIG. 3E, a de-normalized table is shown describing the slot machine player locations, as well as slot machine statistics. For example, at slot machine #1, Jones is the responsible pit boss, current player is ID, the slot machine has been idle for 5 minutes, the payout over time has been $5000, and the number of plays to win on average is 2000.
FIG. 4 illustrates, in accordance with one embodiment of the present invention, a simplified functional diagram displaying event engine 257, as shown in FIG. 3. Event engine 257 is configured to monitor for a variety of events 402 generated throughout the network, as well as within the active data platform itself. A viewset event 406b is a change to a modified data object, such as a database table with a filter. It would be generated by active reports engine 274, as shown in FIG. 2B. A system event 406e is generated from the underlying operating systems or application, for instance, a change in virtual memory or a cache miss. A file event 406f is a change to a given file as stored in a file system. A performance event 406g is a change in a system's performance, such as virtual memory utilization or network latency. A manual event 406h is an event entered by a user. An external event 406i is an externally generated through an API, and is used for such things as functional system verification. A time event 406j is a change in the passage of time, similar to a timer. A date event 406k is a change in date. And, a presence event 462l is the discovery of a client instant messaging device that was previously offline, or the availability status of a recipient has changed.
Event engine 257 is further comprised of four logical components: event generator manager 428, rules manager 408, accountability manager 410, and actions manager 412. Event generator manager 428 maintains the state information of monitored events 402 for rules manager 408. Rules manager 408 is the primary event handler for the active data platform. Accountability manager 410 monitors the delivery status of all reports and alerts. And, actions manager 412 initiates actions based on input from the actions module 434 of the rules manager 408.
Event generator manager 428 maintains state information of monitored events 402 for rules manager 408. Constraint module 426 specifies the parameters that must be satisfied before evaluating an event and/or a condition and/or before applying the rule. Conditions module 424 maintains information for the corresponding conditions module 432 in rules manager 408. It further specifics attributes relevant to the specific event, other than constraints, which must be satisfied before the action associated with the rule may be taken. Events module 425 maintains event information for the corresponding events module 430 in rules manager 408.
Rules manager 408 further includes three components in order to manage active data: an events module 430, a conditions module 432, and an actions module 434. Events module 430 monitors all generated events 402 within the active data platform. Conditions module 432 filters out non-relevant events, and forwards the information to actions module 434, which in turn, initiates new system events based on criteria within rule sets 256. For example, all betting transactions are sent as events to event generator manager 428, which in turn sends the information to the events module 430 of rules manager 408. The conditions module 432 is configured to subsequently filter out all but the blackjack events. The actions module, based on rules 256, then creates an internal system event to instruct the active reports engine 274, as shown in FIG. 2B, to create a report for all blackjack transactions above a threshold, as described in rules 256.
Accountability manager 410 monitors the delivery status of all reports and alerts. It further includes an unaccounted notifies module 438, and a report engine tie module 436. The unaccounted notifies module 438 maintains the status of undelivered alerts and reports. That is, unaccounted notifies module 438 determines if a recipient cannot be reached, subsequently forwards a delivery-incomplete message to rules manager 408. Report engine tie 436 monitors for successfully received reports and alerts, and subsequently forwards a delivery-complete message to rules manager 408. For example, a receipt notification status message is received in accountability manager 410, which is subsequently processed. If the receipt notification status message comprises an acknowledgment, report engine tie module 436 further processes the message and forwards the information to rules manager 408. However, if an acknowledgment is not received, unaccounted notifies module 438 processes the message and forwards the information to rules manager 408, which in turn allows active data cache to escalate alerts and reports to alternate recipients, based on status and presence, should the original recipient be unavailable.
Actions manager 412 initiates events and actions based upon input from the actions module 434. It further includes a notification module 416, a generate event module 414, a run a plan module 418, a web service/MQ module 420, and an others module 422. Notification module 416 forwards any report and alert non-delivery messages to unaccounted notifies module 438. Generate event module 414 generates system events initiated by actions module 434 in rules manager 408. Run a plan module 418 executes script based on rules 256. Web service/MQ module 420 provides an interface to third-party web-services based on WSDL (web-service description language), as well as message queue applications. Others module 422 is an API that allows additional modules to be later added.
FIG. 5 illustrates, in accordance with one embodiment of the present invention, a simplified functional diagram displaying components of rich client 282 and active data cache 256, as shown in FIG. 3.
Rich client 282 is comprised of a main application 536, a rendered reports 540 component, and a re-notify applet 506 component. Main application 536 provides the core environment for client interface to the active data cache 256. It comprises several functional components that can dynamically render incoming data streams from active data cache 256, and locally create reports, such as report builder 538. Rendered reports 540 functions as a viewer to display reports from active data cache 256 which already have been assembled. Re-notify applet 506 displays notification messages to the user.
Line 2: parent.ProcessActiveData(‘’ + viewsetID + “’,’”+strXML + “‘);
Line 3: s1.removeNode(true);
viewsetID indicates the viewset to which the active data applies, whereas strXML refers to an XML document describing new data. Line 1 indicates to the browser that the following section need not be rendered since it is a script section. The execution of line 2 causes the report in the main application (active studio or active viewer) browser to change. The execution of line 3 removes the script from the page. And line 4 indicates to the browser that the script section has ended.
Active data cache further comprises functional components that facilitate the generation of a report. These are: a report execution component 522, a view execution component 524, a report builder component 528, a rule builder component 530, a modifier builder component 532, data object builder component 534, a structured data supplier component 526, and a message center gateway 536.
Report execution 522 creates, caches, and then forwards specific reports based on configuration, to either rendered reports 540 for report builder 538. View execution 524 likewise creates and forwards views to report builder 538. Views are data objects that may be filtered. Report builder 528 pre-processes extracted information from multiple sources, and transforms the aggregated data into a format optimized for use in multiple reports. Rule builder 530 pre-processes stored rules from multiple rule caches for use in the generation of multiple reports. Modifier builder 532 pre-processes intermediate information transformations, such as additional metrics that the user wishes calculated and reported. Data object builder 534 pre-processes data objects into optimized formats for report execution component 522. Structured data supplier 526 forwards external pre-processed data to report builder 528.
FIG. 6 illustrates, in accordance with one embodiment of the present invention, a simplified functional diagram displaying components of message center, as shown in FIG. 3. Message Center 272 provides messaging functionality to the active data platform. It allows reports and alerts to be sent via heterogeneous channels to a plurality of clients. Message center 272 further includes an API layer 605, a presence and notify 612 module, a rich client 614 interface, a thin client 616 interface, an email interface 618, an IM interface 620, other interfaces 624, a local user information 626 module, and an IM client manager 628 module.
API layer 605 provides a uniform application interface for system generated events, such as notification request 608, administration request 606, and presence request 604. Presence and notify component 612 provides the main messaging infrastructure for the active data platform. It insures that reports and alerts are properly routed to intended recipients. Local user information 626 maintains a local cache of user profile information, such as names, passwords, and delivery preferences.
Rich client interface 614 is further coupled to rich client 282, which provides the user with enhanced visualization functionality, such as the capability to graphically render streaming data. Thin client interface 616 is coupled to thin client 112, which further comprises agent application 660. Since thin client 114 is commonly a browser without presence functionality, agent 660 notifies presence and notify component 612 that a given report or alert is being viewed. Agent 660 is commonly an Active X component or a java applet. Email interface 618 is further coupled to email server 632, such as an SMTP server, which forwards messages to email client 115a. IM interface is further coupled to IM client manager 628 which sends a notification request 630 to IM client 115b. Instant messaging (IM) is a means for sending small, simple messages that are delivered immediately to online users. It differs from email primarily in that its primary focus is immediate end-user delivery. Through an IM client application, a user can discover and immediately begin a text conversation, or chat session. Common public instant message systems include AOL Instant Messenger, MSN Messenger, Yahoo! Messenger, Microsoft Exchange 2000 IM, and Lotus Sametime.
For example, a casino's active data cache identifies particularly prolific gaming customer. The active data cache forwards a notification request 608 to message center 272. The presence & notify module 626 then determines if the intended user's device is present, and the user is available. If so, message center 272 forwards an alert through IM interface 620 to IM server 630, and finally to IM client 657.
Unlike like the prior art, the active data platform is general purpose, and is substantially customizable by entering new parameters in active design process 290 and the active studio 291, as shown is FIG. 3A. Furthermore enterprise links 254, also shown in FIG. 3A, allows active data platform to quickly adapt enterprise data sources to changing near-real time BI needs.
Referring now to FIG. 7A, an exemplary active studio screen display is shown according to an embodiment of the invention. This display is used by a user in order to easily manage the generation and secure delivery of data streams, reports, and alerts, without substantial additional computer programming.
The display is divided into three sections, a tab pane 701, a function pane 714, and a display pane 715. Tab pane 701 further allows the user to select one of various functional displays within active studio 304, such as: home 705, my reports 706, shared reports 707, personalize 708, and alerts 709. Home 705 comprises a summarized view of the user's most recent activities. My reports 706 comprises a summarized list of the user's created reports. Shared reports 707 comprises a summarized list of third-part reports to which the user has subscribed. Personalize 708 comprises administrative components that can allow the user to modify report delivery and display parameters. And alerts 709 comprises a summarized list of a user's active alerts.
Function pane 714 displays several functional list boxes within home 705 tab, such as: recent list box 710, new list box 711, list toolbox 712, and actions list box 713. Recent list box 710 displays the user's most recently accessed reports. The new list box 711 displays the user's new reports. The alerts list box 712 displays the user's currently active alerts. And the actions list box 713 allows the user to create and edit reports. Display pane 715 displays the corresponding result of the chose function of function pane 714. In this case, a dynamic financial report comprising a cost bar graph and a percentage change line graph. For example, if a user where to select intraday sales by region in the new list box 711, the corresponding dynamic intraday sales report would be displayed in display pane 715. As sales are booked through out the day, the sales chart would dynamically change as well.
Referring now to FIG. 7B, an exemplary report viewer screen is depicted according to an embodiment of the invention. There are three functions available from this view, launch active studio 719, email 717, and print 718. Display pane 715 allows the user to view the selected report, in this case, a Q1 sales report. Launch active studio 719 allows the user to open the viewed report in the Iteration Active Studio. Email 717 allows the user to mail the viewed report to an email recipient. Print 718 allows the user to print out the viewed report on an attached printer.
Referring now to FIG. 7C, the exemplary active studio screen display of FIG. 7A is shown with the components of my reports tab 706. Function pane 714 further displays two sub tabs, a task bar tab 723 and a folders tab 724. Task bar 723 tab allows the user to select various functions related to the current view, in this case, a report summary view. Folders 724 tab allows the user to organize created or stored reports in an hierarchical format. Two components within the task bar tab 723 are report creation 725 and guide 726. Report creation 725 allows the user to create new active data. Guide 726 provides a source of online help to the user.
Referring now to FIG. 7D, the exemplary active studio screen display of FIG. 7e is shown with additional components of the my reports tab 706, such as the report creation toolbox 723, the report properties toolbox 748, and the guide window 749. The report creation toolbox 723 further comprises functions to easily create specific graphical views 740 of data, based on pre-defined templates. Common views are tables, pie graphs, bar graphs, line graphs, scatter graphs, columnar reports, form reports, geographic (map) views, crosstabs, and Excel integration. These views can be further modified in the report properties toolbox 748. For instance, a user can create an hourly report in which gambling customers are aggregated into different groups based on the value of their accumulated betting that day. The user selects a bar graph for the report, and changes the fill color and font type to the user's preference.
Referring now to FIG. 7E, the exemplary active studio screen display of FIG. 7e is shown with an additional example of the my bar graph 776 function window, which allows the user to further customize the selected report with specific data objects and files. In general, three types of attributes can be customized: views 770, data 772, and properties 774. Views 770 allows a user to modify the types of graphical displays available in which to view the current data, such as a bar graph, line graph, etc. Data 772 allows a user to select the specific data fields to be used in views 770. Properties 774 allows a user to modify other attributes, such as language, number formats, etc. In this case, the user has selected a bar graph template for the report, and has added four fields.
In another aspect of the invention, an Iteration instant messaging & collaboration server is advantageously employed to enable rich secure graphical collaboration in an instant messaging environment. Instant messaging systems commonly monitor the presence of devices, and availability state of users. Presence determines if the IM device is coupled to the given IM network, or at least reachable from the client. Status determines if the intended recipient is available to read and respond to the instant message. Common status categories also include descriptive information, such as “out to lunch”, “busy”, “away from desk”, “important meeting”, etc. For instance, the IM device may be turned on and available, but the user is in an important meeting and does not wish to be disturbed. The user would set the status to “important meeting,” in order to notify other subscribed IM users that he will probably not respond to a text message until after the meeting.
These IM systems, however, commonly lack the capability for graphical peer-to-peer or multi-party collaboration. That is, while short text messages can be sent back and forth in an online chat meeting, there is no comparable means for sharing a graph or report. This would normally have to be compressed and transmitted ahead of time to each recipient, who would each then download, de-compress, and locally store the file. In situations where the underlying data is continuously changing, as temperatures would be in a chemical plant, and impromptu collaboration is desired, pre-mailing a graph would be very impractical, inefficient, and of marginal use to the problem at hand.
Furthermore, public IM systems are not particularly secure. They were originally designed for rapid information text conversations, called chat. IM protocols emerged and were established primarily without supervision from any standard bodies, such as the Internet Engineering Task Force. Subsequently, security was never originally considered and common instant messaging traffic is essentially unencrypted. Just about anyone on the network can secretly monitor a chat conversation.
FIG. 8A illustrates, in accordance with one embodiment of the present invention, an active collaboration server advantageously employed to allow impromptu graphical collaboration between IM users. As shown in FIG. 2B, event engine 257 forwards an event, such as a temperature change, to message center 272, through API 803. Event engine 257 monitors the information in active data cache 256 for pre-determined changes, or events. Message Center 272 provides messaging functionality to the active data platform, allowing reports and alerts to be sent via heterogeneous channels to a plurality of clients.
User manager 811 manages user profile information. Users create profiles based on rules, in which they describe events of which they want to notified. Should such an event occur, user manager 811 notifies message center 272, which in turn contacts the user. Furthermore, user manager 811 selects alternative recipients, if the primary recipient is unavailable, based on user preference and availability status. For instance, user #1 has configured user manager 811 for notification of a specific event, for example a temperature change beyond a certain range. Furthermore, user #1 also wants user #2 to be notified should user #1 be unavailable. Once the event occurs, if user #1 is unavailable, message center 272 attempts to notify user #2. Presence manager 813 monitors for the availability of user devices on the network. For instance, in the previous example, user #1 has a wireless PDA and a SMS phone. If the event occurs, and the PDA is turned-off, an alert is sent to the SMS phone instead.
IIM (Iteration instant messenger) interface 806 provides an interface to the HM client 819-823, an enhanced IM application that allows secure concurrent graphical collaboration and pen-based user input. In general, it is sometimes desirable to implement private instant messaging systems having enhanced capability (including different protocols, encryption capabilities, graphical processing/display capabilities, I/O capabilities, and/or the like) compared to public messaging systems (such as those available from Yahoo™, AOL™, and the like). IIM client is an example of a client in such a private instant messaging system.
IIM client 819-823 commonly connect through active collaboration server 817, which marshals collaboration request from multiple IIM clients, and combines connections whenever possible to improve overall system performance and scalability. Furthermore, active collaboration server 817 synchronizes the text and graphical portions of the IIM session to insure that all users view the same information at substantially the same time.
Active collaboration server 817 further enforces security protocols as implemented by the IIM client, in order to prevent unauthorized interception of the collaboration session. In one embodiment, IIM client 819-823 implements a Triple DES security protocol to protect the collaboration session and insure message integrity. Triple DES is an encryption algorithm whose security is based on a single secret key. Each key is generated (selected at random from all possible keys) for each particular collaboration session.
In another embodiment, only one user in an IIM collaboration session needs authorization to access a given alert or report. By virtue of being authenticated to the session, all other users will receive the alert or report with the security privileges of authorized user, for as long as these users participate in the session.
In addition to IIM interface 806, there is SMTP interface 807 and other IM interface 809. SMTP, or simple mail transfer protocol, is the most common protocol for sending email message on the internet. Through SMTP interface 807, message center 272 can forward email notifications with the event information to email server 815. Unlike IIM client 819-823, however, notification is not automatic. The user must first log onto mail server 815 with a SMTP client 827, such as MS Outlook and download the messages. Likewise, other IM interface 809 represents traditional IM interfaces to currently available IM systems 831, such as AOL Instant Messenger, MSN Messenger, Yahoo! Messenger, Microsoft Exchange 2000 IM, and Lotus Sametime. In contrast to SMTP, notification can be automatic, although communication is generally insecure. Furthermore, graphical collaboration is difficult.
For example, user #1 is a chemical engineer at a petroleum processing plant. He created a profile in user manager 811, in which he wants to be notified of specific events, such as a change in a process temperature beyond a certain range. The event occurs. Message center 272 forwards the notification to active collaboration server 817, through IIM interface 806. Seeing the temperature dynamically change on his device, user # 1 realizes that he wants additional advice from user #2, his manager. Through his IIM device, he instructs message center 272 to initiate a request for collaboration with user #2. Receiving a collaboration request from his subordinate, user #2 accepts the connection. User #1's display is immediately rendered on User 2's device, upon which they collaborate on the matter at hand.
Referring to FIG. 8B, IIM client 823 and active collaboration server 817 as shown in FIG. 8A, are described in greater detail. The user interface of IIM client 823 is comprised of four panes: a graphical display pane 854, a chat text display pane 856, a pen entry pane 858, and a chat text entry pane 860. Graphical display pane 854 is the main rendering window for IIM client 823. It allows a user to make better sense of large amounts of dynamic data, presenting changing information in an easily accessible, graphical visual representation. Chat text display pane 856 displays IM messages in a formatted way, using a choice of colors and styles. Pen entry pane 858, allows the IM device to recognize a user's handwriting. Chat text entry pane 860 allows the user to enter text through a keyboard.
For example, user #1 is a network systems engineer monitoring traffic at a company's e-commerce web site. When network congestion reaches a certain point, active report engine 274 generates a report and forwards it to user #1. Realizing that customers are starting to have difficulty placing online orders, user # 1 initiates a connection with user #2, the web site manager, and user #3, the customer service manager. Once both accept, user #1's display is immediately rendered on the other devices, upon which all can collaborate on the matter at hand.
In another embodiment, an instant messaging manager is advantageously employed to rapidly alert intended recipients of data events. Referring to FIG. 9A, a simplified functional diagram is shown of a rules-based monitoring engine with alert capability, such as in a BI or OLAP application. Enterprise systems 902 comprise the series of applications commonly used to run and manage a business, such as OLTP, data warehouse, and OLAP. OLTP (online transaction processing) applications 902 are those enterprise systems that manage a company's basic transactions, such as order entry and billing systems. In order to conduct meaningful analysis, this information is often further placed in a more stable environment, optimized for random querying, such as a data warehouse. OLAP enables trained users to perform sophisticated ad hoc analysis of data stored in a data warehouse.
Rules-based monitoring engine 916 monitors the applications in enterprise systems 902 for pre-defined events. A rule is a pre-defined procedure comprising the event(s) to be monitored, and the subsequent action or transformation that is required. For example, the event can be a specific transaction, a change in a file size, the publishing of a report, etc. Rules U/I 920 is typically graphical application that allows a user to create sophisticated and complex rules, as well as store the rules in the rules db 918 for use by rules based monitoring engine 916. In this case, the subsequent action is an alert.
When a specific event occurs, rules-based monitoring engine 916 forwards an alert to notification engine 924a, which properly formats the alert and transmits it via email system 915a, pager system 915b, or another system through proprietary API 915c. Alerts sent to email system 915a can further contain an embedded link to a report generated by active reports engine 274, as shown if FIG. 2B.
For example, betting transactions would be normally stored in a casino's game transaction system. Wanting to be extra-attentive to high-rollers, a casino would configure the rules based monitoring engine 916 to observe for any customer whose betting exceeds a specified threshold, upon which an alert would be sent to the closest pit boss with an embedded link to the customer record.
FIG. 9B, in accordance with one embodiment of the present invention, shows the simplified functional diagram of a rules-based monitoring engine with alert capability of FIG. 9A, with the addition of instant messaging manager 924b. Instant messaging (IM) is a means for sending small, simple messages that are delivered immediately to online users. It differs from email primarily in that its primary focus is immediate end-user delivery. Common public instant message systems 915d-f include AOL Instant Messenger, MSN Messenger, Yahoo! Messenger, Microsoft Exchange 2000 IM, and Lotus Sametime. Through a locally installed client application, a user can normally discover the presence of another user, and immediately begin a text conversation, or chat session. Many systems also have the capability for group chat sessions. Instant messaging manager 924b enables the rules based monitoring engine 916 to appear as an IM client to the various instant message systems 915. Therefore no additional software is needed, beyond the IM client application that is more than likely already installed.
As in FIG. 9A, there is shown a simplified functional diagram of a rules-based monitoring engine with alert capability, such as in a BI or OLAP application. Rules-based monitoring engine 916 monitors the applications in enterprise systems 902 for pre-defined events. Rules U/I 920 is typically graphical application that allows a user to create sophisticated and complex rules, as well as store the rules in the rules db 918 for use by rules based monitoring engine 916. In this case, the subsequent action is an alert.
Instant messaging manager 924b properly formats the alert as an instant message and transmits to the intended recipient through one of the various instant message systems 915d-f Unlike notification engine 924a, however, instant messaging manager 924b also monitors the various instant message systems 915d-f for changes in state for any subscribed users. State is comprised of presence and status. Presence determines if the IM device is coupled to the given IM network, or at least reachable from the client. The given instant message system 915d-f normally determines the presence of individual IM clients, and broadcast the information to interested parties, such as the instant messaging manager 924b. Status determines if the intended recipient is able to read and respond to the instant message. Normally whether a recipient is available or not is a true or false question. However, it is common for recipients to customize the unavailable status with further descriptive information, such as “out to lunch”, “busy”, “away from desk”, “important meeting”, etc.
For instance, the IM device can be turned on and available, but the user is in an important meeting and does not wish to be disturbed. The user can set the status to “important meeting” notifying any other subscribed users that he will probably not be able to respond to the instant message until after the meeting. The rules-based monitoring engine 916, realizing that the user will not immediately read the instant message, will choose an alternative present and available user based on a pre-determined notification rule. Rules-based monitoring engine 916 can forward an alert to instant messaging manager 924b for immediate delivery to the intended recipient.
Referring now to FIG. 10A, simplified pseudo-code is illustrated as would be used by IM client manager 628, as shown in FIG. 6, to notifying running processes of user status changes. In procedure 1, ProcessMessagesFromServer with a message argument, IM client manager 628 receives a message from an instant message system. It then determines whether it is a user status in procedure 7, case: UserStatuses. If so, then procedure 15 is called, Process_User_status_messages with an userStatusObjects collection argument. It then updates the user data structure at procedure 18, HashOfUserObjects.update, and subsequently notifies any active processes running on the active server platform of this status change.
FIG. 10B illustrates, in accordance with one embodiment of the present invention, simplified pseudo-code as would be used by IM client manager 628 to determine presence and status for IM clients. In procedure 1, the Boolean IsUserPresent is passed a user (i.e., IM name) and optional degree of presence arguments (i.e., available, non-available, etc.). Initially in procedure 4, if (!HashOfUserObjects.contains) with a user argument, the entered user name is compared to existing list of potential users registered with the active data platform. If not, an exception is thrown, and the procedure returns an error. If true, procedure 8, if (HashOfUserObjects[user].degree>degree) further checks to see if the optional degree of presence variable that was passed to procedure 1 is less than the degree of presence previously stored in the active data platform. The function then returns the result as either a true or false.
FIG. 10C illustrates, in accordance with one embodiment of the present invention, simplified pseudo-code as would be used by IM client manager 628 to subscribe to a specific IM user. IN procedure 1, RegisterInterestInUser( ) is passed a user (i.e., IM name) and optional degree of presence arguments (i.e., available, non-available, etc.). This is passed to procedure 3, CollectionOfInterestingUsers.add, which subsequently registers interest in the user.
In another embodiment, an event engine is advantageously employed to substantially assure the examination of a report by a user. Many analytical products such as business intelligence (BI) and online analytical processing (OLAP) systems are capable of monitoring, scheduling, and broadcasting alerts via email or pager. These systems, however, cannot generally assure that the intended recipient promptly receives and reads the message. This presents a significant problem for time-sensitive information, where minutes or even seconds can make a difference.
In the current invention, the event engine monitors the state of transmitted alerts in an escalation table. Upon receipt of an alert, the user reviews the received report. This is commonly done by manually browsing for the report, or by clicking an embedded URL contained within the alert. If the user does not review the report within a specified time, the escalation table can further escalate the alert to another user based upon predefined rules. For instance, the alert is transmitted to the user's manager.
Referring to FIG. 11, a simplified functional diagram displaying the active data platform is shown with elements of the notification and escalation process. Advantages include substantially improving alert accountability in business intelligence applications. That is, enabling the right person to receive critical information at the right time, and if not, escalating the information to the appropriate person.
Enterprise data 140 comprises those enterprise systems that manage a company's basic transactions, such as order entry and billing systems. Active data cache 256 monitors transaction data packets 1104 generated by enterprise data 140. Active data cache 256 further comprises user data manager 1108, which describes user event interest, and monitoring rule manager 1111, that analyzes data packet 1104 for specific events.
Should an event of interest occur, monitoring rule manager 111 forwards a rule fired message 1116 to the rules list manager 1120 of event engine 257. Rules list manager 1120 further comprises the subsequent actions required for the fired rule, such as the transmission of a notification message to a user, the execution of a subsequent rule, or the generation of an accountability escalation event. In this case, notification message 1130 is forwarded to message center 272. Rules list manager 1120 also places an entry into escalation table 1120, describing notification message 1130, the intended recipient, and the time the message was sent. If a report viewed message 1176 is not received within a pre-determined amount of time, a timeout occurs, and escalation table 1122 notifies escalation manager 1138, which in turn executes pre-configured escalation rules, such as notification of a different person or entity, or the calling of a web service.
Upon receiving notification message 1130, user address manager 1142 attempts to find a device address at which the user is currently available. Message center 272 further embeds a link in notification message 1146, through which the user can review the corresponding report from active reports engine 274.
The user subsequently receives notification message 1146 on his device, in this case, through a browser. If not already viewing the report, the user can either locate the report by manually entering information such as a URL, or can also simply click on an embedded link to the report.
Upon receiving a request, report supplier 1168 of report engine 274, generates the report from information in the active reports table 1170, and forwards it to user device 1150. A report viewed message 1176 is also generated and forwarded to escalation table 1122, which in turn deactivates the previous notification message entry.
For example, a large multi-site manufacturing organization, such as Ford Motor Company, would configure the active data platform to monitor manufacturing processing information, such as increasing part reject rates, or outright machine failures. Should a machine failure actually occur, an alert is forwarded to the appropriate manufacturing process engineer. The alert would contain an embedded link to the specific machine's status report, as generated by active reports engine 274. If the process engineer did not review the report within a few minutes (or any time period specified), a new alert would be immediately forwarded to the local plant manager for immediate action.
Referring to FIG. 12, a simplified diagram of a common instant messaging client displaying notification message 1146 is shown in FIG. 10. In this case, an alert message with a link 1202 to the budgetbuster report is shown. Upon clicking link 1202, a report would open in a new browser window. Furthermore, a report viewed message 1176 would be generated and forwarded to escalation table 1122, as shown in FIG. 10.
In accordance with one aspect of the present invention, rule evaluation is advantageously distributed in order to improve efficiency. To further elaborate, a rule often comprises multiple interdependent components (i.e., event, constraint, condition, and action) that must be evaluated and acted upon.
The first component of a rule is the event, which specifies the event of interest. An example of an event is when the sales volume exceeds 5,000 units. If the sales volume exceeds 5,000 units, the event component is satisfied.
The second component of a rule is the constraint, which specifies the parameters that must be satisfied before evaluating an event and/or a condition and/or before applying the rule. A constraint is usually provided for efficiency reasons and commonly comprises dates, times, and/or frequencies. For instance, if the specific event described in the event clause (e.g., sales volume in the previous example) does not need to be monitored over the weekend or between certain times or more often than a given frequency, then the constraint clause would instruct the event engine to that effect. Typically a constraint is evaluated to be either true or false. In one embodiment, if the constraint is evaluated to be false by the event engine, the satellite evaluation engine is informed accordingly and does not even need to spend its resources evaluating the event. Advantageously, efficiency is improved.
The third component of a rule is the condition, which further specifies other attributes relevant to the specific event, which attributes must be satisfied before the action associated with the rule may be taken. In the previous example, the condition may be to take action only if the sales volume that fires the event is achieved without any special promotion in place. Generally speaking, all conditions may be thought of as events but not all events are conditions.
The fourth component of a rule is the action, which determines the course of action taken should the event, constraint and/or condition components be satisfied. For instance, the sending of a notification to a specific person may be an action that needs to be taken if, as in the previous example, during a weekday (constraint satisfied), the sales volume exceeds 5,000 units (event satisfied) and the sales volume is achieved without any specific promotion in place (condition satisfied).
It should be noted that a rule typically includes an action but a rule does not require all three other components, i.e., event, condition, and constraint. A rule may have one, two, or three of these components as desired by the rule designer. Furthermore, each of the event, condition, constraint, and action components may be a complex Boolean operation comprising multiple sub-components. A preferred embodiment, however, is a single event component per rule.
For example, an event may be satisfied if sales volume is greater than 5,000 and the total revenue is greater than $10,000 (a Boolean operation of two sub-events).
In order to reduce design complexity, a common solution has been to evaluate and execute these rules in a centralized event engine. It is found, however, that the information that needs to be monitored and evaluated can be found anywhere on the enterprise network. It is often the case that the information that must be monitored and evaluated (e.g., the event or condition) may be distributed in one or more machines (such as servers) remote from the computer executing the centralized event engine. In some cases, only 5% of the information may be relevant to a particular rule. Yet when the rule evaluation is centralized, all the data that needs to be monitored and evaluated must be transmitted across the machine and/or process boundaries for the centralized event engine to perform its rule evaluation task. In a networked environment, the remaining 95% of the information that is needlessly sent represents an inefficient use of scarce network resources and bandwidth, as well as a detriment to performance.
In accordance with one embodiment of the invention, rule evaluation is distributed in one or more satellite evaluation engines local to or closer to the origin of the monitored information. The evaluation is distributed in satellite evaluation engines that are in communication with an event engine. Distributed rule evaluation allows the evaluation of a rule to take place with reduced bandwidth usage and delay. Since active data and the timely report/notification responsive to active data are important aspects of the active data platform, minimizing delay is an important consideration, and the distributed evaluation mechanism substantially furthers that goal.
To facilitate the implementation of distributed rule evaluation, it is highly preferable that a universal rule syntax be developed that allows the event engine to analyze any given rule and decides which portion of the rule should be distributed and which portion of the rule should be processed by the centralized event engine.
In accordance with one embodiment of the invention, a rule syntax is developed in which the four components of a rule (event, constraint, condition, action) are specified as clauses in a specific order to facilitate parsing.
For example, the following is a simplified example of a rule that may be executed:
ON viewsset.changed(“QTD Sales”)
IF Time.Between(9,17) AND Date.EndofQuarter( )
Clause 1 describes the event to be executed by the satellite evaluation engine, in this case, the viewset application. The event is a change to QTD Sales information. Clause 2 comprises the constraint clause. It describes that the event will only be monitored from Monday to Friday. Clause 3 comprises the condition, further qualifying the situation that occasions the firing of the action. In this case, if the time was between 9:00 a.m. and 5:00 p.m. (i.e., 1700 hrs), and if the date corresponds to the end of the quarter. Clause 4 describes the action. In this case, the action is to send an email to user named Joe if the event, constraint, and condition are satisfied. Of course, these four components may be ordered in a different order if another syntax construct so dictates.
In accordance with one embodiment of the present invention, a rule is parsed to one or more core events, which are then distributed to one or more appropriate satellite evaluation engines to facilitate evaluation of the distributed event(s). In the previous example, the rule may be parsed and the event pertaining to the sales volume may be distributed to, for example, the satellite evaluation engine associated with the active data cache. The satellite valuation engine then evaluates the monitored data (sales volume in this example) and communicates the result (which may be simply true/false or may include pertinent sales volume data) back to the event engine. Meanwhile, the event engine is offloaded from the task of handling that rule until the event is satisfied.
To further improve efficiency, the centralized event engine may not even parse all components of a rule until the event is satisfied. Thus, if an event is never satisfied, no resource has been wasted parsing the entire rule and preparing the centralized event engine and/or the satellite evaluation engine(s) to handle the other components (which may never need to be handled if the event is never satisfied). Further, the satellite evaluation engine may only send information back to the event engine if the event evaluates to be true. Thus, if the sales volume is below 5,000 in the previous example, no message is sent from the satellite evaluation engine to advantageously avoid wasting bandwidth. Of course, as mentioned earlier, the evaluation of an event may be held off by the satellite evaluation engine to further avoid wasting resource unless the constraint is evaluated to be true by the centralized event monitor.
If the event is evaluated to be true at the satellite evaluation engine (and assuming any applicable constraint is also evaluated to be true), the satellite evaluation engine informs the centralized event engine, which may then evaluate the condition using either the same or another satellite evaluation engine or using the centralized resource. If the same or a different satellite evaluation engine is also used to evaluate a condition, the event monitor may send out the parsed condition and ask for the result of the evaluation of the condition. As the exemplary syntax above indicates, each of the event, constraint, condition, and action may be (but not required in all cases) implemented as a function. Distributing rule evaluation thus may comprise using the appropriate satellite evaluation engine to make the required function call.
Referring to FIG. 13, in accordance with one embodiment of the present invention, a simplified functional diagram illustrating the more relevant components of an exemplary active data platform in which distributed rule evaluation is implemented. Event engine 257 employs a satellite evaluation engine manager 1302 to distribute and manage the evaluation of certain components of a rule in one or more event satellite evaluation engines. The distributed evaluation results are then received by monitor manager 1302 for use by event engine 257.
Satellite evaluation engines 1036a-1036h are some exemplary satellite evaluation engines employed by the active data platform of FIG. 13. Viewset satellite evaluation engine 1306a observes changes to database data objects, such as the modification of a database filter to certain change in the data in the active data cache. Presence satellite evaluation engine 1306b observes changes to the availability of devices and the status of users. Time satellite evaluation engine 1306c observes time changes. Date satellite evaluation engine 1306d observes dates changes. System satellite evaluation engine 1306e observes changes to the underlying operating systems or applications, for instance, a change in the virtual memory or a cache miss. File satellite evaluation engine 1306f observes changes to a given file as stored in a file system. External satellite evaluation engine 1306h observes externally generated events through an API. These external events can be used for integration with other applications. Other satellite evaluation engine 1306i represents additional events that can be programmatically monitored by event engine 257.
In accordance with another embodiment of the present invention, there is provided a universal function call syntax to facilitate the task of distributing certain aspects of rule evaluation to one or more satellite evaluation engines. As discussed above, the rule syntax allows an event, a constraint, a condition, or an action to be specified as a function. Thus, the function call syntax may have the following construct:
- satname.funcname(argument1, argumentN)
where satname is the name of the satellite evaluation engine. Funcname, which is separated from satname by a symbol such as a dot, is the name of the function. The arguments represent optional arguments of the function.
By creating a universal function call syntax, the invention substantially simplifies the task of specifying that a particular satellite evaluation engine should handle a particular event or condition. Furthermore, the function syntax improves scalability in that a new satellite evaluation engine is added to the system, only the satellite evaluation engine name needs to be changed and the remainder of the syntax may be reused. Also, if a new event or condition is specified vis-à-vis an existing satellite evaluation engine, only the function name needs to be changed.
In one embodiment, a rule template library may be provided to allow users to leverage on work done by others. A rule template is simply a rule with certain argument(s) not completely specified. For example, a rule template may specify that if the sales volume exceeds 5,000 on a weekday and no special promotion is involved then notify person X. In this case, person X represents the argument that remains to be filled out to turn the rule template into a rule instance. In this example, the filling out of the person may be automatically accomplished by filling in the missing argument with the identity of the user accessing the rule template. In other cases, the missing arguments may be filled out by the user wishing to employ the rule template.
Of course the user can always edit existing rules (i.e., complete rule instances) to fit his needs. For example, the user can access an existing report and edit the rules therein to modify the arguments related to the event, condition, constraint, and/or action so that the newly created report would contain the information needed. The universal rule syntax renders it simple to implement rule templates and makes it easy for users to convert a rule template to a rule instance or to modify an existing rule to fit his needs.
While this invention has been described in terms of several preferred embodiments, there are alterations, permutations, and equivalents which fall within the scope of this invention. For example, although reference is given to Sagent Business Intelligence Solutions, it should be understood that the invention can also integrate with other business intelligence and ETL solutions. It should also be noted that there are many alternative ways of implementing the apparatuses of the present invention. It is therefore intended that the following appended claims be interpreted as including all such alterations, permutations, and equivalents as fall within the true spirit and scope of the present invention.
Advantages of the invention include optimizing the presence-aware delivery of relevant information to a device in a timely manner, and rendered in a dynamic format. Additional advantages include the continuous extraction of information from data sources without the need for manual queries, or the need of additional user training.
Having disclosed exemplary embodiments and the best mode, modifications and variations may be made to the disclosed embodiments while remaining within the subject and spirit of the invention as defined by the following claims.