SR-D78940 · Issue 542925
Dataflow monitoring enhanced
Resolved in Pega Version 8.3.2
Enhanced monitoring and healthchecks have been added for dataflow and alerts.
SR-D79145 · Issue 543053
Added tokenizer for KeyWord-Based Topic Detection phrases
Resolved in Pega Version 8.3.2
KeyWord-Based Topic Detection was not Working for "Enrollment/Re-Enrollment" as a word, for example a categorization model with the text 'This is an email related to Enrollment/Re-Enrollment'. To facilitate this use, a tokenizer has been added to break should, must, and and words into components for a taxonomy match.
SR-D79909 · Issue 542243
DDS added for automated dataflow run cleanup
Resolved in Pega Version 8.3.2
During a recent upgrade it was seen that there were in excess of 20k dataflow runs, some 2 years old, which slowed down the migration significantly. In order to resolve this, an automated process has been added. This clean-up procedure deletes all the single case, batch, and real time runs older than 30 days which are in the final state - Completed, Completed with Failures, or Failed, and batch and real time runs which are in the Stopped state. The DDS Pega-DecisionEngine.dataflow/run/maxDaysToKeepRuns.should be used configure the retention period. Note that the retention period is calculated since the last processed message and not the creation time of the run.
SR-D82895 · Issue 543605
Restart issue on dataflow service nodes resolved
Resolved in Pega Version 8.3.2
A high percentage of rolling restarts were failing for nodes hosting dataflow services due to prolonged restart times. To resolve this, the system has been updated so Cloud services will not try to recover the node state in case of a restart, since nodeIDs are always new and not reused.
INC-179727 · Issue 704545
Modified batch requestor handling to ensure cleanup
Resolved in Pega Version 8.6.4
A large buildup of batch requestors was seen, and restarting the node did not clear it. Investigation showed that in the case of one class reading a DSS value, a Pega requestor was being created when it was unable to retrieve any Pega context, and this requestor was not cleaned up afterwards. To resolve this, an update has been made to the way the requestor and Pega context is being created along with ensuring it will be cleaned up properly after use.
INC-180246 · Issue 664947
Support for apostrophe added to keyword tokenization
Resolved in Pega Version 8.6.4
A keyword containing an apostrophe was not detected properly in Text extraction model. This has been resolved by updating the annotator used in the tokenization.
INC-192826 · Issue 685797
PMML updated for nested 'if'
Resolved in Pega Version 8.6.4
Running a model on the openscoring API and the Pega PMML model gave different results. This has been resolved by updating the handling for nested 'if' in PMML.
INC-194382 · Issue 697615
ADM Model management page updates
Resolved in Pega Version 8.6.4
When accessing the ADM model management landing page in an environment with a large number of nodes, an "UnresponsiveNodeException" error was thrown from the pylaststrategyresults activity by various nodes on different days. Investigation showed the ADM factory table was truncated, leading to inconsistent ADM data in the Last Responses Section. This has been resolved, and the ADM model management page has been made more resilient to node failures.
INC-196561 · Issue 688846
Natural Language Processing (NLP) topic categorization working consistently for German
Resolved in Pega Version 8.6.4
A few German words were not recognized by the Natural Language Processing (NLP) topic categorization if they were given in capital letters. This was traced to upper case keywords being lowercased during model creation time, but not during prediction time. To resolve this, all words will be lower-cased before the model analyzes the text.
INC-206109 · Issue 706238
Delayed Learning works correctly with Volume Constraints
Resolved in Pega Version 8.6.4
When running an outbound campaign using NBA Designers predictions and delayed learning in conjunction with volume constraints, the adaptive models ended up with incorrect feedback if the prediction had an Alternative label configured with a defined waiting window. This has been resolved with the addition of a new flag 'isVolumeConstraintEnabled' during dataflow creation which will check for volume constraint, and a filter in HandleResponses for that flag.