An updated BPM Analytics technology preview is available to use with the latest IBM Business Process Manager. The latest version only works with V8.6 release with CF 2017.12 (December 2017).

Overview

The key value of IBM Business Process Manager (BPM) is in streamlining custom enterprise business processes to better optimize service and cost. It does this namely through 1) custom process applications to manage work, and 2) process analytics for workers, managers, and analysts to assist their decision making in the execution and management of work and the design of processes. This technology demonstration aims to enhance BPM capabilities for (2) process analytics, with modern technologies which can excel in the era of big data and analytics, taking advantage of the full potential of the rich information into business operations afforded by custom process apps executing in BPM.

BPM Analytics aims to enhance BPM for two scenarios: 1) process analytics – providing enhanced process analytics features directly within the BPM offering targeting BPM user roles, and 2) 3rd Party analytics – providing enhanced features to publish process data to external data and analytics solutions provided by IBM, customers, and partners.

This technology preview enhances the existing BPM Dynamic Event Framework (DEF) to publish business events in JSON message format, and after a series of the reformatting and aggregation, store the query-optimized documents at the Elasticsearch.  Once the query-optimized documents are stored in Elasticsearch, business users can then build dashboard with the Kibana tool.  In addition, this update also brings performance enhancements, adds the ability to send messages through Kafka, as well as backups and restores business events to and from OpenStack Swift Object Storage system.

Elasticsearch and Kibana are parts of the popular Open Source Elastic Stack (http://www.elastic.co). Elasticsearch is a distributed, search and analytics engine. Kibana is a dashboard that allows you to visualize your Elasticsearch data, and perform drill down analysis.



Reference Topology



Sample Kibana dashboard of team’s performance

This technology preview is available for download on the IBM BPM Github (https://github.com/ibmbpm/bpm_analytics), and is to be used the latest IBM Business Process Manager with V8.6 only.  To find more details, check out my InterConnect presentation on this topic.

This technology preview provides no explicit support, guarantee or warranties by IBM.

To collaborate with other users of this technology demonstration post your question to dwAnswers . To make it easier for community members to find your question, we recommend tagging your questions with BPM and BPM Analytics.

14 comments on"Updated! Technology Preview: BPM Analytics"

  1. Elias Teodoro March 31, 2017

    How can I to connect Process Performance Data Base from IBM BPM to Kibana?

    • Allen Chan March 31, 2017

      Hi Ellas, that is not yet supported. Replaying data in PDW database is something we are considering to support in the future.

  2. […] BPM Analytics capabilities empower managers, business analysts, and process owners to improve business operations […]

  3. […] BPM Analytics capabilities empower managers, business analysts, and process owners to improve business operations […]

  4. Will there be an impact on the performance of IBM BPM instance if we configure BPMEventEmitter to provide each transaction information. Do you think there should be native support in a way choice to store event details into NOSQL rather than RDBMS which is current configuration.

    • Allen Chan May 19, 2017

      Depending on the number of events you’re emitted, there will be performance overhead. The emitted events are currently stored as JSON in Elasticsearch in this technology demonstration.

  5. Filip Lou May 16, 2017

    Hi, I have had a lot of success hooking up BPM to Watson Analytics. I was wondering if that option has been explored as well.
    The simplicity and flexibility of creating insights in the data is very appealing to business users.

    • Allen Chan May 19, 2017

      Using Watson Analytics is an option – and this technology demonstration does not conflict with the use of Watson Analytics. One of our goals is to get the business events out of BPM, so customers can apply their own analytics software on them.

  6. There is a difference between the Lombardi-event (mon:monitorEvent) and BPEL-events (wbi:event).
    BPMEventEmitter.war uses the xmlToJSON method in JSONUtils.class (OICommon-1.0.0.jar lib).
    public static JSONObject xmlToJSON(String xml)
    {
    JSONObject monEvent = XML.toJSONObject(xml).getJSONObject(“mon:monitorEvent”);
    return monEvent;
    }
    So when you try to emit the BPEL-event with the BPMEventEmitter you get this error:
    Caused by: com.ibm.bpm.mon.oi.json.JSONException: [mon:monitorEvent] is not existed. BPMEventEmitter.war does not support BPEL (BPM Advanced, Integration Designer) DEF events?

    • Allen Chan May 19, 2017

      Supporting BPEL events is an important requirement but it is currently not supported in this technology demonstration. As a workaround, you can write your own emitter but some amount of custom work would be required to aggregate the events for consumption in Kibana.

  7. Can I connect and use an external JMS such as TIBCO EMS to emit the events?

    • In theory, if you can connect TIBCO EMS to WAS as a JMS destination, it will work. On the other hand, we have not tested it with TIBCO.

  8. What kind of data can you get out of the Event Emitter, is it just the tracked field and the business data? Or can we even get the Complex Business Object that are user-defined?
    In addition, is there a way to connect elastic/kibana to the bpm rest api ? If yes please do advise on how to do so.

  9. Can I connect elastic/kibana to IBM BPM Rest API in order to get the data? or can we track complex data objects that are user defined in order to create dashboards/reports out of them?

Join The Discussion

Your email address will not be published. Required fields are marked *