If using the dropcolumn filter during extraction, in conjunction with the Batch Applier (eg Replicating to Redshift, Hadoop, Vertica) writes would fail with a CSV mismatch error due to gaps in the THL Index.

However, for JDBC appliers, the gaps are required to ensure the correct column mapping

To handle the two different requirements, a new property has been added to the filter to control whether or not to leave the THL index untouched (the default) or to re-order the Index ID's

If applying to Batch targets, then the following property should be added to your configuration. The property is not required for JDBC targets.

--property=replicator.filter.dropcolumn.fillGaps=true

Issues: CT-1025

Fixed an issue that would prevent reading remote binary logs when using SSL.

Issues: CT-958

Restored previously-removed log file symbolic links under $CONTINUENT_ROOT/service_logs/

Issues: CT-1026

Fixed a bug where tpm diag would generate an empty zip file if the hostnames contain hyphens (-) or periods (.)

Issues: CT-1032

Improve ability to find needed binaries for commands: tungsten_find_position, tungsten_find_seqno and tungsten_get_rtt

Issues: CT-1054