APAR status
Closed as documentation error.
Error description
In rare cases customers may experience repeated restarts of Integrated Synchronization which may suspend the subscription. Subsequently, the subscription will be reported as in ERROR. DSNX881I messages with the following content may occur even if the replication has been running successfully before: message: "Dumped error log stream to file 2020-04-23T06:32:48.555_dump_1.db2zlog" time: "2020-04-23T06:32:48.559Z" severity: WARNING originator: "ApplyServerInstance" message: "Thread LogParserThread threw an exception which was not caught" time: "2020-04-23T06:32:48.559Z" severity: ERROR originator: "TerminatingUncaughtExceptionHandler" In the trace the following error string my be shown: 2020-04-23 06:33:25 SEVERE dwa.replication.apply.utilities.TerminatingUncaughtExceptionHand ler uncaughtException: Thread LogParserThread threw an exception which was not caught java.lang.AssertionError: invalid column length read from the column offset arrays: 0. Length must not exceed column precision: 128 The issue affects IDAA V7 maintenance levels 7.5.0.x and 7.5.1. It occurs when the value of a key column has a length of 0. Integrated Synchronization (InSync) will enter a loop and finally suspend; the steps of the loop are: 1) Failure showing the messages listed above 2) Restart InSync process 3) Failure with the same messages 4) After four retries to restart InSync it will be finally suspendend. Customer will not be able to get InSync out of this state again and needs IBM support to clear the situation. A fix will be delivered with maintenance level 7.5.2. Contact IBM support to clean up the system and apply a configuration that avoids the occurrence of this is issue. Additional Keywords: InSync TS003626660 RTC#24734
Local fix
To prevent the occurrence of the issue, avoid replication of tables with empty values in the Key columns, for example: avoid "" as value of a varchar key column. If the issue already occurred, proceed as follows to get the subscription up and running again: 1.DisableReplicationForDatabaseSystem / EnableReplicationForDatabaseSystem (if reloads are acceptable). However, the issue may reoccur if rows with empty Key columns are replicated. 2. Disable affected tables: 2a Determine which tables are affected by the issue (this can be done by checking replicated-tables.json along with the dumped .dbz2log). 2b Set affected tables to state ERROR in replicated-tables.json 2c Remove suspension marker (file name SUSPENDED). 2d If you have removed all affected tables, replication will continue for the tables not affected.
Problem summary
The corresponding program code responsible for handling of key columns was corrected accordingly. The fix is part of maintenance level 7.5.2.
Problem conclusion
Upgrade the accelerator maintenance level to 7.5.2 or higher.
Temporary fix
Comments
APAR Information
APAR number
PH25111
Reported component name
ANYTCS ACCLTR Z
Reported component ID
5697DA700
Reported release
750
Status
CLOSED DOC
PE
NoPE
HIPER
NoHIPER
Special Attention
NoSpecatt / Xsystem
Submitted date
2020-05-06
Closed date
2020-10-16
Last modified date
2020-10-16
APAR is sysrouted FROM one or more of the following:
APAR is sysrouted TO one or more of the following:
Fix information
Applicable component levels
[{"Business Unit":{"code":"BU054","label":"Systems w\/TPS"},"Product":{"code":"SG19M","label":"APARs - z\/OS environment"},"Platform":[{"code":"PF054","label":"z\/OS"}],"Version":"750"}]
Document Information
Modified date:
19 October 2020