SR-D66325 · Issue 527661
Data pages explicitly cleared after QP use
Resolved in Pega Version 8.2.6
The Util Node was showing as Offline in the Search Landing Page, and when Jobs were submitted for execution from other Nodes the message "Detected active run with unreachable nodes" was logged. The util node, configured as a backgroundprocessing node, was running QPs; the queue size for custom QPs is 500 messages /queue items per minute, but investigation showed the requestor level and thread level data pages corresponding to the QP activities were not being cleared after use. This led to high heap memory issues that made the node unreachable, and has been resolved by adding code to explicitly remove the data pages when processing has finished.
SR-D66615 · Issue 536241
API added to minimize calls to set parameters in datapage
Resolved in Pega Version 8.2.6
Differences were seen after upgrade when using a datapage associated with a property which was application-specific and that had Data Access for the property defined as “Copy data from a data page" with parameters whose value was set by an activity from the associated property of type single page. Prior to upgrade, only one call was seen to set the values for the parameters in datapage, but post-upgrade multiple calls were used to set the values to these parameters. This was traced to logic used for extraction and invoking the perform, and has been resolved by adding the new API 'DeclarativeAction.shouldExtractWatchedProperties'. It will return true or false, depending whether the extraction of watch properties is needed or not.
SR-D66986 · Issue 539413
BIX command line support improved
Resolved in Pega Version 8.2.6
Post-upgrade, it was seen that changes to the engine which validate that Data-Admin-DB-Name is not defined multiple times caused the BIX command line extracts to no longer work the same way. In order to improve backwards compatibility, updates have been made to loosen the validation in JdbcConfigurationFactory to always prefer prconfig settings if there are multiple definitions.
SR-D67034 · Issue 536806
Modifications made to reduce IncrementalIndexer disk space usage
Resolved in Pega Version 8.2.6
Queue Items were being added back to the queue when the ElasticSearch cluster health was not good, causing pyFTSIncrementalIndexer data to constantly grow and consume disk space while processing the same records over and over. As an immediate remedy, the scheduler time has been reduced from 1 hour to 10 minutes to stop the processing of queue Items. The issue will be further evaluated for better possible solutions in a future release.
SR-D67232 · Issue 530008
Email listener updated to handle encoded MIME attachment
Resolved in Pega Version 8.2.6
When processing an email with an attachment that was itself an encoded MIME part, the email listener generated a NullPointerException around an extraction error. Analysis showed that a content-type misled the system into thinking an attachment was another message part. To resolve this, EmailMessageParser.java has been modified to avoid attempting to process an attachment as a MessagePart if it cannot get the content of the part as a Message.
SR-D68204 · Issue 530034
Connect-FTP now s3 repository aware
Resolved in Pega Version 8.2.6
In order to support using the Connect-FTP feature with the s3 repository as the source path, the system has been updated to check the local path provided for a filespec and attempt to retrieve the file for the Connect FTP call from there. This uses the "local file" field , using the format of "file://<storageDestinationAlias>/targetfolder" .
SR-D69836 · Issue 529728
Context retained during nested automation
Resolved in Pega Version 8.2.6
Issues were seen with context being passed correctly with automation. When a Data Page that was used to abstract the call to a REST API used a Data Transform which in turn used a Data Page which also used a call to a REST API to retrieve an OAuth Token, an error appeared. The same worked when a URL was provided on the Connect REST. When a service case was launched but a Flow Action wasn't submitted, then another service case was added and submitted, the post-processing robotic automation activity used the run robotic automation activity from the previous service case instead of the current service case targeted for completion. This has been resolved by updating the webwb • pzpega_ui_doc_submit.js file to pass the correct MDIContext based on the event target of FA submission while invoking RDA automation.
SR-D69927 · Issue 537479
Cleanup added for staging directory
Resolved in Pega Version 8.2.6
Temporary files from imports and exports (from DevOps) were filling up the staging area disk space because there was no automatic process for cleaning up these local files. This has been resolved by adding an enhancement that will clear the directory on Engine Startup and any time ParUtils.setStagingDirectory gets called to initialize the staging directory.
SR-D71041 · Issue 537895
Default compiler pool size increased
Resolved in Pega Version 8.2.6
An issue was seen where the nodes were continuously terminating and Stream service was not started. Error messages indicated a "WARN - Long running request detected for requestor", and that tasks were waiting for the generic object pool. Analysis showed that it was possible for a deadlock to form during the startup of the system or during the first few rule executions and compilations when there were more than five rules trying to compile. Once the variable LibraryMetadata.TypeEnvironment.resolver (ClassLoader) is initialized, the problem would resolve. To prevent the deadlock, the default compiler pool size has been increased from 5 to 10.
SR-D71977 · Issue 535601
Obj-Browse will retrieve more than 50 records
Resolved in Pega Version 8.2.6
After performing multiple skims, a newly added property was not retrieved when an Obj-Browse was performed. In Obj-Browse action, when a class is mapped to an external table and does not contain BLOB column pzpvstream, in a specific case where pxObjClass value is a reference to a parameter, a list of properties mapped to the class explicitly by querying pr4_rule_property table using NativeSQL is fetched. This call by default returns only 50 records: in cases where there are more than 50 properties for an external class, some of the properties from the obj-browse response mapping are therefore missed. To resolve this, the limit of 50 records has been removed from this NativeSQL call.