CUBRID 11.2 Patch 9 has been released.

 

In this release, issues found after 11.2.8 released have been fixed.

 

1. Improvement and Fix

  - Fixed the "Cannot allocate query entry any more" error in Java SP

  - Fixed the issue where the vsnprintf function was not working properly in the WINDOWS version

  - Fixed an issue where unloaddb would cause an infinite loop when data exceeded 1MB in a JSON type column 

  - Fixed an issue where NULL values remained in the db_authorization system table after deleting a user

  - Fixed an issue where rownum was not changed to orderby_num in the ORDER BY removal logic

  - Improved recovery logic up to a certain point (restoredb)

  - Fixed an error that occured in the unloaddb/loaddb utility when the current_val and max_val values of serial are the same

  - Fixed an error message that occured when incorrect information is entered in ha_node_list and ha_replica_list

  - Improved string handling to support multiple encodings

  - Fixed the sort-limit optimization to work even when using bind variables

  - Corrected incorrect trace information in UNION, DIFFERENCE, INTERSECTION, and CTE

  - Resolved permission issues when performing 'for update' statements on system tables

  - Modified the behavior of vacuum_er_log_error() in the internal vacuum functions

  - Modified to prioritize the creation of the java directory when running a Java SP server

  - Fixed an issue where localhost could not be found when setting certain environments in ha_make_slave.sh

  - Fixed an issue where incorrect results were produced when a join relationship was not established

  - Improved the error message regarding the 'reverse unique index' comment when performing unloaddb

  - Fixed an issue when there was a space between the class name and parentheses in a Java SP declaration

  - Modified the broker/cas processes created by cub_manager to prevent them from becoming zombie processes

  - Removed unnecessary directories from the release file

  - Fixed an issue where NULL values remained in the db_auth system table after deleting a user

  - Modified the show statement to only display views owned by the connected account

  - Modified the method for registering a server in dblink and locating the server when the user schema is omitted in the query

  - Fixed a Java SP related issue when executing SELECT statement with a 'WHERE 0 <> 0' clause

  - Fixed an error that occured when using user variables in TIME_FORMAT() function

  - Modified the results of to_char(), time_format(), and date_format() functions to express TZM information identically

  - Improved the performance of strcat(), an internal function used in time_format() and date_format()

  - When using the oracle_style_empty_string value, the result of the replace(?,'-','') function is modified to be NULL

  - Fixed an issue where NULL was incorrectly handled in the '||' operation when using the oracle_style_empty_string value

  - Simplified repeated function calls and condition checks in char_isspace() and char_islower()

  - Resolved a compiler error where numeric values were incorrectly generated as double values when creating Java SP results

  

 

  Please refer to the 11.2.9 change logs(https://github.com/CUBRID/cubrid/releases/tag/v11.2.9) for details.

  

2. Precautions

  - When using heterogeneous(Oracle/MySQL) DBLINK, you must use ODBC Drvier for Unicode of Oracle/MySQL.

 

3. Environmental Requirements

  - Linux: 64bit, kernel 2.4 or higher, Windows: Windows7 or higher (Windows 7 requires the latest security updates)

  - JRE 1.8 or higher is required when using JSP or CUBRID Admin/CUBRID Manager

 

4. Download

  - Product download : https://www.cubrid.org/downloads

  - Product source code : https://github.com/CUBRID/cubrid/tree/v11.2.9