Yes. The ingstop command is issued when the machine is shut down.
Use kanjieuc for Ingres non-Unicode data types (for example char and varchar). If you want to use the Ingres Unicode data types, you must convert your UTF-8 data to UTF-16 format for storage in Ingres (regardless of the Ingres character set setting). To convert to UTF-16 format, you can use, for example, iconv(3).
On Linux systems, the file /bin/sh is a link to a shell such as bash, ash, ksh, or zsh. This shell is invoked when a Bourne shell script is run. Ingres was developed and tested on a Linux system using GNU bash, version 2.05b-50. Limited, successful testing has also been done with the ksh and zsh shells.
Make sure that you are not running the "createdb" program that is provided by PostgreSQL. Make sure that the PATH setting for the shell from which you install and start Ingres includes Ingres executable directories before other executable directories.
The TERM_INGRES environmental variable defines the type of terminal you are using, so that the function keys will work when using the forms-based tools, such as CBF or QBF. On Linux, setting TERM_INGRES to konsolel will work for most terminals.
During installation, the value of TERM_INGRES is set to konsolel. You can change this value if it does not meet your needs.
Running Ingres in an xterm, set TERM to xterm and set TERM_INGRES to vt100fx. Then use xmodmap to determine and set your function keys. To show current settings, use the command xmodmap –pke . (To determine the syntax for your version of xmodmap, use man xmodmap.)
You will probably find that there are no bindings for KP_F1 through KP_F4; you will need to bind them.
For example: to bind keys Shift+F1 through Shift+F4, create a file "mykeys" that contains:
keycode 67 = F1 KP_F1
keycode 68 = F2 KP_F2
keycode 69 = F3 KP_F3
keycode 70 = F4 KP_F4
Then issue the command:
xmodmap mykeys
Shift+F1 through Shift+F4 will now be defined as PFK1 through PFK4.
No. Standard kernels and kernels compiled with default values (without modifying the Linux source headers) should provide adequate resources. For additional information, see the Readme file.
You might, however, need to increase the maximum allowable size for shared memory segments. You can do this by running /sbin/sysctl –w kernel.shmmax=<new value> as root. Change is immediate and does not require a reboot.
Ingres will fail to start if any of the kernel parameters do not meet required values. You can check these parameters using the syscheck utility in $II_SYSTEM/ingres/utility, as follows:
syscheck
If syscheck reports any potential problems, use the following to generate a list of suitable parameters:
syscheck –c
The output is in the format expected by /sbin/sysctl. If the output is written to a file, the new values can be applied as follows:
syscheck –c > out.file
/sbin/sysctl –p outfile
Ingres was compiled using the GCC compiler version 3.3.3 with the following switches:
-fsigned-char -fno-strength-reduce -D_REENTRANT -DLINUX -D_GNU_SOURCE -DXLIB_ILLEGAL_ACCESS -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE -fPIC -O
No. While this step is required on some UNIX systems, it is not required for this version of Ingres for Linux.
The g77 compiler (which is bundled with many, if not all, Linux distributions) does not support some of the statements that the ESQLF pre-compiler generates. More information is available at gnu.org.
If your terminal is using the same character set as Ingres, you should not have a problem. If characters are not displaying correctly, try starting the terminal with a specific character set.
For example, if you are using SHIFTJIS (Japanese Double Byte character set) as the II_CHARSETXX setting in a kterm, start the kterm with the following command:
kterm -km sjis
If you encounter problems using double byte character sets and kterm, try using the rxvt terminal instead. For example, to start the rxvt terminal with the SHIFTJIS character set, issue this command:
rxvt -km sjis