CUBRID 11.2 Patch 2 has been released.
In this release, issues found after 11.2.1 released have been fixed.
1. Improvement and Fix
* Query related improvement and bug fixed
- Improved performance when the query through sort-limit optimization when using bind variables in limit syntax.
- Improved performance when the query through the index scan when using the stored function in the where clause.
- Fixed a segment fault that occurred when using built-in functions using call syntax. (eg CALL TO_DATE('12/25/2008'))
- Fixed a segment fault that occurred when performing the index scan query after creating a function index using Mod() function.
- Fixed a segment fault that occurred when query using Hash List Scan is stopped by killtran.
* user schema & synonym related improvement and bug fixed
- Improved to omit the for syntax when performing alter synonym comment and fixed to become NULL if the comment is an empty string.
- Fixed an error in which the owner name is missed in the table name in the SQL file generated when performing the synccolldb.
- Fixed an error that the owner name is missing on the table name in the SQL file generated when performing the Synccolldb.
- Fixed an error that occurred when loading an object file separated by the user name and a table name (eg %id [u2].[t1])
- Fixed to separate user_name and table_name in unloaddb's object file. (eg %id [u2].[t1])
* dblink related bug fixed
- Fixed the schema file of unloaddb to have the create server syntax located before the view syntax with dblink.
- Improved the dblink related string processing method of oracle/mysql. (Fix errors that occur when patching Korean columns in ksc5601db)
* javaSP related improvement and bug fixed
- Improved to change the communication method of the Javasp server by setting java_stored_process_uds. (yes : use unix domain socket, no : use tcp socket) (Note : It can only be set in Linux OS, and Windows OS operates only as tcp socket.)
- Fixed an error in which UnsatisfiedLinkError occurs in JavaSP class using JNI (so or dll).
- Fixed the JavaSP server start error when setting up CUBRID_TMP environment variable. (Note : CUBRID_TMP environment variable can only be set in Linux OS)
- Fixed a NullPointerException error that occurred when a string column of 255 bytes or more is passed as a JavaSP argument or exist in the select list with JavaSP.
- Fixed an error in which the JavaSP server does not end when the Heartbeat ends in the HA environment.
* Etc
- Fixed an error in which transactions after the recovery point are also recovered when point-in-time recovery is performed using restoredb -d. (Note: In order to maintain compatibility, if a point-in-time recovery is performed using a previous version of the backup file in this version, transactions after that point-in-time may be recovered together)
- Fixed an installation error in Prolinux8, Oracle Linux8
Please refer to the 11.2.2 change logs(https://github.com/CUBRID/cubrid/releases/tag/v11.2.2) 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/v11.2.2