SR-D91894 · Issue 557220
Class Loader cache values wrapped with WeakReference for improved cleanup
Resolved in Pega Version 8.2.7
Out Of Memory exceptions due to Metaspace were observed on Web tier instances running with a max Metaspace size of 2GB. Investigation showed that in PRClassLoaderDB, mLoaderCache values were wrapped with a SoftReference. To resolve this, the Class Loader cache values will be wrapped with a WeakReference, allowing the cache to free memory when it is no longer needed. JVM options has also been added to switch types of References being used if there is a preference.
SR-D92707 · Issue 551694
QP exception handling improved
Resolved in Pega Version 8.2.7
There is currently a configurable maximum size limit for QP items of 5MB. If the message size exceeded 5 MB, the message failed to be enqueued to Kafka, but still ended up in the delayed message table where it remained. This caused issues with pzDelayedQueueProcessorSchedule JS as a result. To resolve this, the system has been updated to better detect Kafka errors related with message size and move any corrupted item to the broken message queue with the appropriate message attached.
SR-D94019 · Issue 553783
Alerting features added for Listeners
Resolved in Pega Version 8.2.7
An enhancement has been added for Email Listener alerts. At the end of each listener's processing cycle, the system will determine the average processing time per message, the total processing time of the listener, the total unread messages count, and total number of messages processed. In each AESRemote agent run, this data will be published to the PDC, where it will be analyzed and any necessary alerts raised.
INC-134932 · Issue 578544
Timeout for node to join cluster increased and made configurable
Resolved in Pega Version 8.5.1
Hazelcast timeouts were causing application restart failures. This was traced to a rare case scenario of heavy processing on the primary node causing the new node to not be able to join the cluster within the designated 5 minute threshold. To compensate for this, the default time to join has been increased to 9 minutes and a prconfig setting has been added to configure the time setting.
INC-133270 · Issue 581839
Automatic agent restart added for temporary connection issues
Resolved in Pega Version 8.5.1
An agent that was stopped due to an issue with the database connection did not recover when the connection was reestablished. To resolve this, updates have been made that will handle an automatic restart of the agents when there are temporary connection issues.
INC-135359 · Issue 587038
CORS header configuration added for Stream tier
Resolved in Pega Version 8.5.1
An enhancement has been added that allows setting the CORS header (Access-Control-Allow-Origin) to allow *.DOMAIN.com on Stream tier (/stream/view & /stream/click) calls by way of the setting /stream/httpHeaders .
INC-134069 · Issue 583024
Added Kafka logging
Resolved in Pega Version 8.5.1
In order to help diagnose issues encountered during Kafka background processing, additional logging has been added to QueueProcessorDatasetHandler.
INC-142792 · Issue 594499
Modified thread data page cleanup
Resolved in Pega Version 8.5.1
Some cases were becoming stuck at a particular stage and not proceeding further even though the task in the backend was complete. Investigation traced this to work done that removed declarative pages at the thread level to optimize performance while other clipboard pages were retained, and the issue has been resolved by deleting all thread data pages after every item processed by the agent.
INC-125122 · Issue 581255
JMX authentication added
Resolved in Pega Version 8.5.1
Support has been added for JMX authentication through user and password.
INC-130886 · Issue 590374
Bad Stream error after passivation resolved
Resolved in Pega Version 8.5.1
A "Bad Stream" error was seen when attempting to reassign a case to the Workbasket, and a save of the work object failed. This was caused by a corruption of the work object during passivation, and has been resolved.