Contents
TIBCO® Data Streams 10.6.3 adds the following updates and new features:
- TIBCO® Data Streams Now Supports Windows Server 2019
-
Windows Server 2019 is now a supported operating system for development and deployment.
- TIBCO® Data Streams Now Supports AlmaLinux OS 8
-
AlmaLinux is now a supported operating system for deployment, and is the recommended migration path for applications deployed on CentOS Linux 8.
- TIBCO® Data Streams Now Supports Debian Linux 11
-
Debian Linux is now a supported operating system for deployment.
- TIBCO® Data Streams Now Bundles Java 11.0.15.1
-
On all supported operating systems, the TIBCO Data Streams installation kit provides the runtime components of Oracle JDK version 11.0.15.1 in a subdirectory of the StreamBase installation directory, for private use to run StreamBase program fragments and applications from within StreamBase Studio. As before, StreamBase detects and prefers a system-installed JDK, and uses the bundled JDK as a fallback. See further details about the JDK requirements of this release in the JDK entry of the Supported Configurations page.
The following features are deprecated as of StreamBase 10.6.3.
- CentOS Linux 8 In Container and Machine Image Builds
-
Support for CentOS Linux 8 is still in effect for Streaming 10.6. However, as Red Hat has declared that CentOS Linux 8 has reached end of life, there will be no security updates. Container or machine builds that relied on CentOS package repositories, such as Docker or Ansible, will not work and must be migrated to other compatible Streaming-supported Linux distributions.
For Streaming, the options are Red Hat Enterprise Linux, Oracle Linux, and AlmaLinux. AlmaLinux is the recommended option for container builds like Docker; use the image name
almalinux:8.
This section provides a list of errors corrected in release 10.6.3 of the TIBCO StreamBase® component of TIBCO® Data Streams.
| Fixed in 10.6.3 | |
|---|---|
| Number | Resolution |
| TIBCO StreamBase® 10.6.3 incorporates all fixes resolved in the release 10.0 series through 10.6.2 and all fixes resolved in the release 7.7 series through 7.7.8. | |
| SB-50392 | The JMS and or EMS connect adapter could introduce a deadlock to an application in certain configurations if a connect command tuple was sent to the connect adapter while some other adapter in the same container was trying to connect to the same configured JMS/EMS server. |
| SB-50260 SB-50261 | If the XML to Tuple Operator receives invalid XML strings, no tuple was emitting on the status port. |
| SB-51443 | If the JMS Consumer adapter received the new message within a small interval of the JMS Commit adapter in a transacted session, it could not emit additional messages. |
| SB-51159 | Changes to role assignments would not take affect until the AMS server was restarted. This issue is now resolved, as cache invalidations happen every 2 minutes. |
| SB-50141 | If encrypted values were added to any HOCON configuration file, user could no longer launch fragments, but can run applications only. |
| SB-51813 SB-51816 | The source Distributed Router now receives the notification that the target node has become available and delivers the queued tuples for that node as well as reporting the change from the Status output port. No queued tuples are dropped. |
| SB-50028 |
An error occurred when using SOL_CLIENT_ACKNOWLEDGE as a value to the JMS/EMS
adapters' acknowledge-mode configuration setting.
The JMS/EMS adapter now supports the SOL_CLIENT_ACKNOWLEDGE acknowledge mode.
|
| SB-51317 | In the previous release, OPC UA adapter was not accepting Encrypted credentials. |
| SB-51259 |
There was an issue when trying to cleanly shut down a node containing a Kafka Consumer that was unable to connect to its specified Kafka server. |
| SB-51090 | Users were not able to load decision table using ADS when running in K8s. |
| SB-51237 | Users were not able to configure SB request/reply adapter operator with custom message converter dynamically. |
| SB-51478 | Users earlier faced error on memory table due to refactoring. |
| SB-51892 | Users saw some lag in the price captured by the CEP engine during Market open and market close for some tickers due to obsolete BLPAPI version. |
| SB-51672 | The parameters pages in the documentation contained outdated SB7-style configuration settings. |
| SB-51446 | The Database Change Data Capture tried to start CDC on SQL server Table in spite of the option being disabled. |
| SB-50667 | Users were not able to reconnect to JBDC datasource and face some unexpected error. The issue had been addressed in this release and now the application tries to reconnect to the database after disconnect. |
| SB-51855 SB-51854 | The OSI PI Event Frames adapter earlier did not return attribute value if frame was not closed. This issue is fixed in this release. |
| SB-49544 | The add container command ignored the --modulesearchpaths and --resourcesearchpaths parameters specified by the user. |
| SB-51860 | Add AllDescendants=false to your query string to prevent the OSI PI adapter from traversing child frames and emitting the children of Event Frames returned by the query. |
| SB-50971 | If a module reference was running in a parallel region and an error occurred, the runtime transaction was not rolled back. |
| SB-51540 | In a few cases, StreamBase Studio was launching the LiveView demo client on old port 10080 instead of the new default port 11080. |
| SB-49285 | An Aggregate operator running in a parallel region did not close its window if the output tuple reached an Error Output Stream within that region. |
| SB-51607 | The Database Change Data Capture (CDC) adapter was not using the JDBC Data Source HOCON configuration to enable pooling options. Because of this, CDC Adapter could not monitor multiple CDC-enabled tables efficiently. |
| SB-32314 | Initialization of a Router Operator on one node could deadlock indefinitely at startup, if the target Router in another node was configured to not start automatically, otherwise prevented from initializing. |