CUBRID 11.2 Patch 4 has been released.
In this release, issues found after 11.2.3 released have been fixed.
1. Improvement and Fix
- If you cannot use the ping command in the HA environment, it provides a alternative method. (see the setting value ha_tcp_ping_hosts)
- Modified so that 0.0.0.0 cannot be used in Ha_Ping_hosts in HA environment.
- In the HA environment, when loading a large amount of data using Loaddb -C, fixed an error that are not replicated to Slave.(Note: When using Loaddb -C, the table with FK must first load the table that the FK references. According to the order of loading, it may occur that cannot be replicated to slave.)
- Fixed an error that do not apply to the value of the backup_dest_path variable in the ha_make_slavedb.sh script.
- Error correction of duplicate output of query results using DB_VCLASS(VIEW).(Note: The VIEW query spec in the system catalog has been modified, which may occur the same error when using the previous version of DB)
- Improved that column types set to AUTO-INCREMENT properties cannot be changed to a type that does not support AUTO-INCREMENT.
- Fixed a segment fault that occurred when using the dblink() function in a DELETE/UPDATE syntax.
- Fixed a segment fault that occurred when using the alter server.
- Fixed an error that changed the result due to the limit clause when using USE_DESC_IDX hint.
- Modified to remove the existing default value when changing the data type to an ALTER COLUMN.
- Fixed an error in which the ROWNUM value is incorrectly output by View merging in a specific environment.
- Fixed an error that do not work correctly by the Hash list scan when VOBJECT is included in the predicate of the query.
- Fixed a segment fault that occurred when using Covered index scan in a CONNECT BY query.
- Fixed an error of -1042(concat string size max size) that occurred when performing a string concat query in a specific environment.
- Fixed an error caused by the failure to remove the predicate '1=1' generated by constant folding during MRO(Multiple Key Ranges Optimization).
- Fixed a segment fault that occurred when using a specific built-in function in the CONNECT BY clause.
- Fixed an error in which the error of the query used in JavaSP did not pass.
- Upgraded to the modified version of the fd leak of the junixsocket library used by Javasp.
- Fixed an error that occurred when the total size of bind factors exceeded 4096 bytes when using server-side JDBC(default connection) in javasp.
- When using loadjava -y, it has been modified to operate with dynamic loading.(A UnsatisfiedLinkError error may occur using jni in javasp during dynamic loading. cubrid javasp restart is required if this error occurs.)
- Fixed an error that returned the wrong results by the incorrect cache initialization in Javasp.
- Fixed an error in which the CURSOR returned by javasp in a Windows environment dose not work normal.
- Improves the processing of absolute paths in linux format among set values in a Windows environment.
- Fixed an error that did not generate demodb during installation in a Windows environment.
- Fixed an error that do not progress on reinstallation after installation interruption in a window environment.
- Fixed a segment fault caused by invalid tde_default_algorithm setting.
- Modified the manager server to process the information in statdump to 64bits.
- Fixed a segment fault that occurred during Unloaddb in the ubuntu-based OS.
- Fixed a segment fault that occurred while reading LOG_REDO data when using Flashback.
Please refer to the 11.2.4 change logs(https://github.com/CUBRID/cubrid/releases/tag/v11.2.4) for details.
- 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
- Product download : https://www.cubrid.org/downloads
- Product source code : https://github.com/CUBRID/cubrid/tree/v11.2.4