Db2 for IBM i agent changelog

Arena batches are handed over in the order they were written instead of being re-sorted by transaction id, which on Oracle LogMiner inverted two changes sharing an SCN and wrote the older value to the target;

GSSD-1370: CDC on a logical file is refused again, instead of starting without errors and never reading anything because its rows are journaled under the physical file’s name;

Booleans and maps keep their type through the arena cache, instead of reaching the target as "true"/"false" and as a non-JSON string;

The change readers are stopped when their agent is unassigned or closed, and the MySql and MariaDB binlog reconnect loop is cancelled on stop;

Added the source lag probe, reporting the unread change backlog and how far behind the last read transaction is;

GSSD-1357: The two unchecked arena length fields are now bounded, so a damaged arena is reported as a corrupted record instead of surfacing as a NumberFormatException or advancing the recovery scan into garbage;

GSSD-1357: An arena is no longer mapped at a size the filesystem refused to give it, which on a full disk wrote into a sparse region with no blocks behind it;

Replaced the Chronicle Queue source cache with ArenaCache, the local cache first proven on Oracle LogMiner. See Source agent caching layer;

GSSD-1324: Entities whose columns declare a position their table disagrees with are refused, and column positions are read from the database instead of from what was stored;

GSSD-1290: Target columns are resolved by name alone, with the positional fallback reported;

GSSD-1290: A column no field answers to is rejected instead of being read by position;

Per-table read CCSID override for character columns;

GSSD-1316: Improve journal reader repositioning logic on transient failure;

Target character columns: a value longer than its target column is now refused instead of being stored truncated with the transaction reported as written. Affects the IBM i series, Informix, MS SQL Server and Oracle targets, on the direct and the bulk write path alike (GSSD-1292, GSSD-1267);

GSSD-1290 IBM i (AS/400) source: fixed CDC stopping with "Missing journal definition for column <name>" when two entities replicate the same table;

Fixed journal deserialization of decimal columns by resolving the packed/zoned converter from the QDBRTVFD physical field type;

GSSD-1267: an undecodable journal entry no longer stalls the whole journal read, behind the skipJournalEntriesWithDeserializationErrors agent option, disabled by default;

Fixed journal deserialization of decimal columns by resolving the packed/zoned converter from the QDBRTVFD physical field type;

GSSD-1267: an undecodable journal entry no longer stalls the whole journal read, behind the skipJournalEntriesWithDeserializationErrors agent option, disabled by default;

GSSD-1270: Fixed receiver rotation detection when the receiver name contains digits;

GSSD-1255: Fixed tables not quoted on AS400 source;

Added support to override column defined ccsid with a custom one;

GSSD-1248: Removed tables count from schemas to improve loading performances;

Fixed commitment control error on IBM i using Bulk;

GSSD-1189: Fixed high cpu usage due to AS400 JDBC connection lifetime default value too high (now set to 60 minutes, configurable in the agents settings);

GSSD-1189: Fixed wrong comparison logic in TransactionID that can cause high CPU usage;

Changed logic to check connection using a dummy query in case of failure of direct connection check;

GSSD-1176: Fixed management of AS400 journal receiver name rotation (9999 → 0000);

GSSD-1160: Added null bytes replicated as null from AS400;

GSSD-1162: Added flag to enable or disable multi-member support !!!!!! WARNING !!!!!!!! The flag is disabled by default. If you are already using the multi-member feature, please enable the flag in the agent configuration under advanced settings enabling property “Enable multi-member physical file support”;

GSSD-1120: Fixed deduplication logic for primary keys columns with case insensitive collation;

Fixed an issue where DELETE operations were treated as UPDATE operations when the agent was used as a target;

Added support for multimember physical files: the agent can now discover and replicate data from all members within a multimember physical file, treating each member as its own entity with independent CDC tracking via journal APIs, snapshots, and bulk loads;

Added log entry to trace table’s reorganize with related notification;

Sped up agent’s warm up cycle, by…​ a lot;

Added support for checkpointed dedicated journal reader;

GSSD-1117: Fixed column nullability always true;

GSSD-1074: Fixed lob type management and logs improvements;

Fixed journal receiver rotation mechanism;

GSSD-1064-63-62-61-60-55-53-43-42: Fixed an issue that was causing checkpoint not to advance when fetched rows where all marked to be skipped;

GSSD-1040: Fixed journal data reader rotation when there are more than 300 entities belonging to a single journal to overcome per journal-filter API limits;

Added precision fallback for null datetime precision;

Implemented teardown logic for component setups to speed-up entity remove flow;

Added logic to recreate staging table for bulk operations to keep it aligned with entity schema;

Fixed user name detection under native journal APIs;

GSSD-805: Added support for Bulk loads over tables without Journal;

GSSD-977: Added journal polling configuration property;

GSSD-971: Added socket timeout configuration property for AS400ConnectionPool;

GSSD-975: improved journaled tables management;

Improved support for enhaced columns discovery (precision, lenght, scale);

Fixed an issue that was bothering users when Journal is enabled on tables after an entity was already configured (it will now attempt to refresh its own metadata first);

GSSD-890: Fixed an issue with VARGRAPHIC data type support;

Fixed possible duplicate key error using bulk operations for target agent.

Fixed date shifting during snapshot;

Fixed a synchronization stall caused by incorrect cache index iteration.

Snapshot performance enhancements;

GSSD-825: Fixed a journal reading issue that could cause checkpoint progress to stall;

GSSD-795 & GSSD-825: Fixed an issue that could cause the agent to miss data changes when using cached journal mode and in few other specific edge scenarios;

Performance & stability improvements;

GSSD-808: Fixed a calendar date conversion issue when reading from the journal date values like "0001-01-01";

GSSD-793 & GSSD-800: Fixed an issue causing the agent to throw an exception saying "Receiver not found" when failing to establish comunication with the CL program used for receivers retrieval;

New journal reader mode: we have now fully transitioned toward the Journal APIs;

GSSD-764: Given compatibility with large BLOB and CLOB data types;

GSSD-686: Fixed historical timezone offsets (e.g., Uruguay UTC-3:30) causing unwanted date/time shifts behaviors;

GSSD-760: Fix wrong bulk staging table queries with tables with only PKs;

Performance improvements thanks to an improved display journal strategy;

Improved binary decoding performance with reduced compute resources usage;

Introduced a new configuration option to enable/disable the use automatic conversion of EBCDIC encoded columns to UTF-8 when reading data from the source, also known as translate binary=true;

GSSD-744: Fixed Columns definitions fetch;

GSSD-685: Support for RRN as a pivot column in logical partitioning;

Fixed exception cast failure;

Fixed bulk logic;

GSSD-670: Snapshot performance improvement;

Fixed an issue when using the dedicated journal reader function together with table containing special characters;

Replace drop of staging table with Truncate operation in case of Bulk loads usage;

Agent compatibility upgrade to Gluesync 2.2;