Installation Guide: OpenLink ODBC to JDBC Bridge Connector (Driver) for Linux (Enterprise Edition)

OpenLink · March 13, 2025, 11:19pm

This installation guide walks you through installation of the Enterprise Edition ODBC to JDBC Bridge Driver to a Personal or Application Server host functioning as a client to a database that’s accessible via its native JDBC Driver.

Step 1: Download the Installer Archive

Visit the OpenLink ODBC Enterprise Edition Driver Download Page to download the appropriate ODBC to JDBC Bridge Driver.

Alternatively, curl can be used to download the installer archives and install.sh script directly:

curl -O https://download3.openlinksw.com/uda/components/misc/install.sh
curl -O https://download3.openlinksw.com/uda/components/9.0/x86_64-generic-linux-glibc25-64/oplodbc.taz
curl -O https://download3.openlinksw.com/uda/components/9.0/x86_64-generic-linux-glibc25-64/jdbc_sv.taz
curl -O https://download3.openlinksw.com/uda/components/9.0/x86_64-generic-linux-glibc25-64/oplrqb.taz   
curl -O https://download3.openlinksw.com/uda/components/9.0/x86_64-generic-linux-glibc25-64/odbc_admin.taz

Step 2: Pre-Installation Configuration

A 64-bit JVM must be installed and configured.
A third-party or native JDBC drivers that connects to the target database must be available
The JDBC Driver must be installed on the same machine as the Multi-Tier “Enterprise” Edition Request Broker and Bridge Agent(s) for JDBC Data Sources.
The JDBC driver jar file (or a symbolic link to it) must be included in the active CLASSPATH environment variable.
The JDBC driver class name must be known.
The JDBC connection URL that connects to your target JDBC Data Source must be known.
The location of the libjvm.so shared library using by the OpenLink JDBC Bridge agent must be set in the LD_LIBRARY_PATH environment variable
Generally, the latest JVM from http://java.com/ is the best choice.

Step 3: Installation

Log onto the target machine for the installation and create a suitable OpenLink installation directory {OPENLINK_INSTALL}.
Copy all the downloaded components to the {OPENLINK_INSTALL} directory, which are the install.sh, oplrqb.taz, oplodbc.taz, odbc_admin.taz, jdbc_sv.taz files.
Install OpenLink components using the install.sh shell script with the command –

 sh install.sh

Take care to enter correct information when prompted pertaining to ports, passwords, etc and ensure that you note these down for future use.
The installation should finish without error.

Step 4: Configuration

To setup the OpenLink environment, from the {OPENLINK_INSTALL} run the command –

. ./openlink.sh

Open the file {OPENLINK_INSTALL}/bin/oplrqb.ini with a suitable text editor such as vi.
Locate the [Environment JDBC18] section for setting the driver specific JDBC configuration and edit to reflect your environment, setting the CLASSPATH to the location of the JDBC Driver jar file and the LD_LIBRARY_PATH to the location of the libjvm.so shared library determined in Step 2 –

[Environment JDBC18]
CLASSPATH          = {JDBC Driver JAR file}
LD_LIBRARY_PATH = {Directory of libjvm.so file}
;; PATCHNULLSIZEOFSQLCHAR = 4096
;; SQL_DBMS_NAME = hugh

Place the License file(s) (oplrqb.lic, jdbc.lic) either purchased or obtain during the evaluation download process in the {OPENLINK_INSTALL}/bin directory.
Start the OpenLink License Manage (oplmgr) by going to the {OPENLINK_INSTALL}/bin directory and running the command –

./oplmgr +start

Start the OpenLink Request Broker (oplrqb) by going to the {OPENLINK_INSTALL}/bin directory and running the command –

./oplrqb -v     ;;; To start in background mode
or
./oplrqb -fd    ;;; To start in foreground debug mode

Step 5: Create Data Source Name (DSN)

Open the file {OPENLINK_INSTALL}/bin/odbc.ini with a suitable text editor such as vi.
Locate and edit the sample OpenLink Multi-Tier ODBC data source (DSN) called [OpenLink] created during the installation process and edit as follows –

[OpenLink]
Driver          = {OPENLINK_INSTALL}/lib/oplodbc.so
ServerType      = JDBC18
Host            = localhost:5000   ;;; replace "localhost" with the hostname the OpenLink Broker (oplrqb) runs on if not on the same machine
Database        = {JDBC Driver Class Name}
Options         = {JDBC connection URL}
Username        = {username}
Password        = {optional password}
FetchBufferSize = 99
ReadOnly        = no
DeferLongFetch  = no
JetFix          = no
WideAsUTF16 = Y
Description     = Sample Informix Multi-Tier Connection

Step 6: Test connection

Using the iODBC Runtime Environment

Use the OpenLink iODBC iodbctest sample tool packaged with the HTTP-based OpenLink ODBC Administrator as follows to make a test connection –

$ cd {OPENLINK_INSTALL}
$. ./openLink.sh 
$ iodbctest "DSN=OpenLink;UID={username};PWD=xxxxxxxx"
iODBC Demonstration program
This program shows an interactive SQL processor
Driver Manager: 03.52.0507.0105
Driver: 08.02.1217 OpenLink Generic ODBC Driver (oplodbc.so) 

SQL>

SQL statements can now be issued directly against the target database from the SQL> prompt –

SQL> select count(*) from orders

(count(*))       
-----------------
23               

 result set 1 returned 1 rows.

SQL>

The ODBC to JDBC Bridge Driver is now ready for use with target ODBC application(s).

Using the unixODBC Runtime Environment

When using the unixODBC Driver Manager the WideAsUTF16 = Y parameter needs to be added to the target ODBC DSN, if not in place already, to enable successful connections with applications built using the unixODBC SDK, e.g., its isql test application –

$ isql OpenLink {username} {password}
+---------------------------------------+
| Connected!                            |
|                                       |
| sql-statement                         |
| help [tablename]                      |
| echo [string]                         |
| quit                                  |
|                                       |
+---------------------------------------+
SQL> select count(*) from orders              
+------------------+
| (count(*))       |
+------------------+
| 23               |
+------------------+
SQLRowCount returns -1
1 rows fetched
SQL>