V4.1.46 - connect_cdc_sqdata - Latest

Connect CDC (SQData) Release Notes

Product type
Software
Portfolio
Integrate
Product family
Connect
Product
Connect > Connect CDC (SQData)
Version
Latest
ft:locale
en-US
Product name
Connect CDC (SQData)
ft:title
Connect CDC (SQData) Release Notes
Copyright
2026
First publish date
2018
ft:lastEdition
2026-03-10
ft:lastPublication
2026-03-10T11:41:18.328000
L1_Product_Gateway
Integrate
L2_Product_Segment
Data Integration
L3_Product_Brand
Precisely Connect
L4_Investment_Segment
Application Data Integration
L5_Product_Group
ADI - Connect
L6_Product_Name
Connect CDC
  • Apply (SQDATA and SQDENG) and Replicator (SQDRPL) Engines:
    • Fix memory leaks in both Engines when writing AVRO formatted data.

      Engines writing AVRO formatted Kafka at V4.1.44 or V4.1.45 using the REPLICATE script command are particularly susceptible to impact and should be immediately upgraded to version 4.1.46 to avoid significant impact.

    • Support mutual (mTLS) authentication to Confluent registry using new sqdata_curl.conf file containing SSL certificate key pair. At this time, the file must be located in the Engine's working directory.

    • Issue error message but ignore Kafka fatal error -187 and -193.
      Note: This is a temporary accommodation for customers who do not want an engine to fail immediately as a result of certain Kafka fatal errors, specifically, -187 ALL BROKERS DOWN, and -193 RESOLVE. Some may expect the situation to resolve itself in short order and would rather have the engine remain active until it does. However, we cannot know if it ever will.
  • Parser (SQDPARSE) and Apply (SQDENG/SQDATA) Engine:
    • Hide passwords specified on the Engine Script RDBMS statement from display in listings and logs.

  • Replicator (SQDRPL) Engine:
    • Fix error preventing shutdown that caused the replicator to remain active after closing its controller socket on sqdmon stop command.

    • Fix a serialization error in the scheduler thread that could result in a hang.
    • Add a health checker for the reader thread to detect certain resource shortages or extended lock wait time.

    • Add Replicator --reader-timeout=<seconds> command line argument.
      Specify the number of seconds after which the replicator will terminate due to resource shortages or extended lock time. The default is 600 seconds (10 minutes). The applicable conditions are:
      • No free uow structures. Stopped due to MAX IN FLIGHT and none of the in-flight UOWs completed during the timeout interval.
      • No free cache storage. Stopped due to exhausting CACHE MEMORY and no memory was freed during the timeout interval.
      • Extended serial lock wait. Stopped for a worker to process a serial uow and the worker made no progress during the timeout interval.
  • IMS Unload Utility (SQDIMSU):

  • Controller Daemon (SQDAEMON):

    Stop sending unintended and harmless SQDAEMON message header to engines.

  • DB2 Capture (SQDDB2C):

    Prevent DBCLOB "unsupported type" error from causing capture termination when a table has a DBCLOB column type. DBCLOB columns are not published to engines.

  • z/OS Capture and Publisher hot standby tasks:

    Issue z/OS capture and publisher hot standby task messages as Write to Operator (WTO). SQDC067W and SQDC068I messages, previously issued only to the SYSPRINT DD in hot standby capture/publisher started tasks are now additionally issued as WTO to permit interception by automated tools.

  • z/OS RACF and USS installation:

    Revise jobs to support automated implementation and eliminate manual steps. Some of the z/OS installation sample jobs in the CNTL distribution library were modified to eliminate manual steps in the process. Modifications were made to ALLOCZDR, ALLOCZFS, CRDAEMON, RACFZFS. New members ALLOCOWN and SUSHELL were added.