SR-B70652 · Issue 325760
Read operations updated for Datastax 3.1.x use
Resolved in Pega Version 7.3.1
In the 3.1.x Datastax driver reads are no longer retried by default. Therefore, the read operations have now been explicitly marked as idempotent to force the Datastax driver to retry timed out reads.
SR-B70652 · Issue 325924
Timeout error resolved for write future
Resolved in Pega Version 7.3.1
With retries enabled, the write future was timing out before retries could be completed. This has been fixed by removing the timeout on the write future since timeouts will be caught using Datastax driver exceptions.
SR-B74934 · Issue 327017
Facebook connector refined to handle page deletion while the AADrivetech is active
Resolved in Pega Version 7.3.1
A Facebook connector trying to connect to a invalid page was causing out of memory errors. In this scenario, a Facebook Page that was no longer required was deleted but the AADrivetech Facebook connector was left running. This connector, in its attempts to contact the Facebook page and retrieve the getFrom parameter which was now null, started to consume resources on the node leading to the outage. To solve this, the code has been modified to not expect the user name to always come from Facebook but instead put in an anonymous name if the page disappears.
SR-B76526 · Issue 326533
Backwards compatibility enhancement for @when() validation
Resolved in Pega Version 7.3.1
After upgrade, a Strategy rule which referred a 'when' with the syntax @when(isOnlineApplication) was failing with design time validation saying that when rule was not found in the SR and instead it had to find it in the Customer/Applies to class. For the @when() issue, the root cause was that the validation context of @when() was switched from Applies-To class to the Step Page class due to a change in the core engine. The behavior of expression parsing for when rule calls was changed in release 7.2. It used to take the Apply-to class to validate the existence of the rule, but not it is taking the Step Page class. And Strategy rule doesn't push/pop stackframe due to performance reasons, thus StepPage on stack for Strategy is always the same as Primary page. For greater compatibility, the system will set the PageContextClass to the Apply To class so the expression parser can validate the setup which is expected at run-time.
SR-B69409 · Issue 317548
CLOB type handling enhanced in Marketing
Resolved in Pega Version 7.3.1
Pega Marketing uses the out-of-the-box DSM activity pxRunDDFWithProgressPage for triggering Data flow runtime. After upgrade in DB2 environments, the DF execution completed but the activity returned a failure. It was found that if a CLOB column is present in the filter of a NativeSQL object invoked from an activity, the status had an error message even though functionality was fine. To ensure consistent behavior, CLOB has been added to supported data types so the severity of the exception can be changed so as to not show error message in activity's status.
INC-128342 · Issue 594185
Improved cleanup for adm_response_meta_info
Resolved in Pega Version 8.6
The adm_commitlog.adm_response_meta_info column family was growing, leading to gradual increase in CPU utilization on the Adaptive Decision Management (ADM) nodes over time. Investigation showed that the compaction on the adm_response_meta_info table was not being triggered by the ADM service, and the compaction did not remove rows that belonged to models that had been deleted. To resolve this, compaction of the adm_response_meta_info table has been moved from the ADM client nodes to the ADM service nodes, which will correctly trigger the compaction on a predefined schedule. The compaction logic has also been refactored to remove rows that belong to models that have been deleted.
INC-136634 · Issue 591679
Thread contention resolved for InteractionHistoryQuery
Resolved in Pega Version 8.6
High thread contention was observed while running Inbound load tests. This was traced to pzLoadInteractionHistory invoking a generated strategy to retrieve ihFields from InteractionHistoryQuery using the getUsedInteractionHistoryFields() method: because this method is synchronized, it caused a bottleneck. To resolve this, the synchronized methods have been replaced with the appropriate data structures.
INC-136969 · Issue 585547
Section and Paragraph rule types added to Revision Manager
Resolved in Pega Version 8.6
An enhancement has been made to add support for section and paragraph rule types in revision management. With this change, the sections and paragraph rule types can be added to overlay and can be added to change requests and modified.
INC-138037 · Issue 586595
Strategy handling updated for very large systems using IH summary
Resolved in Pega Version 8.6
When a Strategy in a Real-time dataflow used IH Summary on a system with more than 5000 groups for one eventKey, the message "Error retrieving aggregates from Cassandra KVS" intermittently appeared. Investigation showed that if the number of result rows was greater than the FETCH_SIZE (set to 5000), it meant another read to Cassandra was required and an exception was generated. To resolve this, updates have been made so that instead of returning maps, the system will return iterators and change them to map on the calling thread.
INC-138103 · Issue 585640
Enhancement added for node heartbeat recovery process
Resolved in Pega Version 8.6
Nodes were not showing up in the admin portal even though they were up and running and could be seen in the pr_sys_statusnodes table. The exception "An exception was encountered while invoking the cluster membership listener callback" was seen. All nodes became visible again after multiple restarts. The root cause was traced back to a temporary database connectivity problem. The database itself was fine according to database monitoring reports, but a network problem, a slow database query, or another issue prevented Pega from establishing a connection for more than a minute. An enhancement has been made to resolve this: if a node becomes unhealthy due to the service registry missing due to a failed heartbeat, the heartbeat will try to recover after 60 seconds and keep trying every 30 seconds until it succeeds.