Published Release Notes

The UpdateAdaptiveModels agent causes an exception after Pega 7.2 to 7.2.1 upgrade

After the Pega 7 Platform is upgraded from version 7.2 to 7.2.1, the log files might show an error that is caused by the UpdateAdaptiveModels agent. This agent is enabled by default and is responsible for updating scoring models in the Pega 7 Platform. If you use adaptive models in your solution, you can avoid this error by configuring the Adaptive Decision Manager service. If you do not use adaptive models, disable the UpdateAdaptiveModels agent.

For more information, see Configuring the Adaptive Decision Manager service and Pega-DecisionEngine agents.

Remove obsolete versions of the PredictiveAnalytics-UI and PredictiveAnalaytics-IntSvcs rulesets

After you upgrade the Pega 7 Platform to version 7.2.2, remove the PredictiveAnalytics-UI and PredictiveAnalytics-IntSvcs rulesets from your application as their versions are obsolete. Beginning with Pega 7.2.2, use the Pega-DecisionScience ruleset that contains the latest versions of the PredictiveAnalytics-UI and PredictiveAnalytics-IntSvcs rulesets.

For more information, see About ruleset versions.

Reconfiguration of the Adaptive Decision Manager service after upgrade to Pega 7.2.1

After you upgrade the Pega 7 Platform to version 7.2.1, you need to reconfigure the Adaptive Decision Manager service. Beginning with Pega 7.2.1, the Adaptive Decision Management (ADM) service is native to the Pega 7 Platform and is supported by the Decision data node infrastructure.

For more information, see Services landing page.

Reconfiguration of the Adaptive Decision Manager service after upgrade to Pega 7.2.2

After you upgrade the Pega 7 Platform to version 7.2.2 from a version prior to 7.2,1, you need to configure the Adaptive Decision Manager service. Beginning with Pega 7.2.1, the Adaptive Decision Management (ADM) service is native to the Pega 7 Platform and is supported by the Decision data node infrastructure.

For more information, see Services landing page.

Interactions in flows are no longer supported by the Run Interaction shape

The Run Interaction shape in flows has been replaced by the Run Data Flow shape, which supports running a single case data flow with a strategy. Flows that include the Run Interaction shape continue to work; however, you must now use the Utility shape to reference any new interactions that you create.

For more information, see Running a decision strategy from a flow and About Interaction rules.

Cassandra 3.11.3 support for Pega Platform

Increase your system's reliability and reduce its memory footprint by upgrading the internal Cassandra database to version 3.11.3.

For on-premises Pega Platform™ users, after you upgrade to Pega 8.3, it is recommended that you manually upgrade to Cassandra 3.11.3. You can upgrade to Cassandra 3.11.3 on all operating systems except IBM AIX. If you do not want to upgrade to Cassandra 3.11.3, you can continue to use Cassandra 2.1.20, which is still supported.

For Pega Cloud Services 2.13 and later versions, Cassandra automatically upgrades to version 3.11.3 after your environment is upgraded to Pega Platform 8.3.

For information on how to manually upgrade to Cassandra 3.11.3, see the Pega Platform 8.3 Upgrade Guide for your server and database at Deploy Pega Platform.

Upgrade impact

After an on-premises Pega Platform upgrade, you still have the older version of Cassandra and must manually upgrade.

What steps are required to update the application to be compatible with this change?

To upgrade Cassandra, you must create a prconfig setting or a dynamic system setting with the new Cassandra version and then do a rolling restart of all the Decision Data Store nodes to upgrade them to the latest version of Cassandra.

Text analytics models editing and versioning

Pega Platform™ now supports editing and updating training data for text analytics models.

Pega Platform also supports the versioning of text analytics models. When you update the model, Prediction Studio creates an updated model version. You can then switch between the model versions.

Upgrade impact

In versions of Pega Platform earlier than 8.3, the training data for text models was stored in the database. In Pega Platform version 8.3 and later, the training data for text models is stored in Pega Repository. You cannot build new models without setting the repository. After the repository is set, all text models are automatically upgraded and will work normally.

What steps are required to update the application to be compatible with this change?

After a successful upgrade, set the repository in Prediction Studio before building or updating any Natural Language Processing (NLP) models.  In Prediction Studio, click Settings > Text Model Data Repository.

For more information, see:

• Increase the accuracy of text analytics models by adding feedback data (8.3)
• Updating training data for text analytics models

Text analytics models migration

Pega Platform™ now supports the exporting and importing of text analytics models. For example, you can export a model to a production system so that it can gather feedback data. You can then update the model with the collected feedback data to increase the model's accuracy.

Upgrade impact

In versions of Pega Platform earlier than 8.3, the training data for text models was stored in the database. In Pega Platform version 8.3 and later, the training data for text models is stored in Pega Repository. You cannot build new models without setting the repository. After the repository is set, all text models are automatically upgraded and will work normally.

What steps are required to update the application to be compatible with this change?

After a successful upgrade, set the repository in Prediction Studio before building or updating any Natural Language Processing (NLP) models.  In Prediction Studio, click Settings > Text Model Data Repository.

For more information, see:

• Increase the accuracy of text analytics models by adding feedback data (8.3)
• Exporting text analytics models
• Importing text analytics models

Changes to the architecture of the Data Flow service

In Pega Platform™ 8.4, the architecture of batch and real-time data flows uses improved node handling to increase the stability of data flow runs. As a result, there are fewer interactions with the database and between the nodes, resulting in increased resilience of the Data Flow service.

If you upgrade from a previous version of Pega Plaftorm, see the following list for an overview of the changes in the behavior of the Data Flow service compared to previous versions:

• Responsiveness
• Updates to lifecycle actions
• Starting a run
• Triggering pre- and post-activities
• Selecting a node fail policy
• No service nodes and active runs

Responsiveness

Nodes no longer communicate and trigger each other, but run periodic tasks instead. As such, triggering a new run does not cause the service nodes to immediately start the run. Instead, the run starts a few seconds later. The same applies to user actions such as stopping, starting, and updating the run. The system also processes topology changes as periodic tasks, so it might take a few minutes for new nodes to join runs, or for partitions to redistribute when a node leaves a run.

Updates to lifecycle actions

To make lifecycle actions more intuitive, the Stop action consolidates both the Stop and Pause actions. The Start action consolidates both the Resume and Start actions.

You can resume or restart stopped and failed runs with the Start and Restart actions. The Start action is only available for resumable runs and continues the run from where it stopped. The Restart action causes the run to process from the beginning. Completed runs can only be restarted. If a run completes with failures, you can restart it from the beginning, or process only the errors by using the Reprocess failures action.

Starting a run

New data flow runs have the Initializing status, and start automatically. You no longer need to manually start a new run, so the New status is now removed.

If there are no nodes available to process a run, the run gets the Queued status and waits for an available node.

Triggering pre- and post-activities

The system now triggers pre-activities on a random service node, rather than on the node that triggered the run.

The system triggers post-activities only for runs that complete, fail, or complete with failures. If you manually stop a run with the Stop action, the post-activity does not trigger. However, restarting the run with the Restart action triggers first the post-activity, and then the pre-activity.

You can no longer choose to run pre- and post-activities on all nodes.

Selecting a node fail policy

For resumable runs, you can no longer select a node fail policy. If a node fails, the partitions assigned to that node automatically continue the run on different nodes.

For non-resumable runs, you can choose to restart the partitions assigned to the failed node on different nodes, or to fail the partitions assigned to the failed node.

No service nodes and active runs

If the last data flow node for an in-progress run fails, the run remains in the In Progress state, even if no processing takes place. This behavior results from the fact that data flow architecture now prevents unrelated nodes from affecting runs.

Limits on active data flow runs

You can now configure a maximum number of concurrent active data flow runs for a node type. Set limits to ensure that you do not run out of system resources and that you have a reasonable processing throughput. If a limit is reached, the system queues subsequent runs and waits for active runs to stop or finish before queued runs can be initiated, starting with the oldest.

For more information see, Limit the number of active runs in data flow services (8.5).

Upgrade impact

If you have many data flow runs active at the same time, you might notice that some of the runs are queued and waiting to be executed.

What steps are required to update the application to be compatible with this change?

You do not have to take any action. After the active runs stop or finish, the queued runs start automatically. The default limits are intended to protect your system resources, and you should not see a negative impact on the processing of data flows. However, if you want to allow a greater number of active data flow runs to be active at the same time, you can change the limits. For more information, see Limiting active data flow runs.